CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
E
xternal current account deficits or
surpluses in some major economic
areas—notably the United States and
Asia—have reached record-high levels,
and expectations are that they will stay large or
increase for some time. Many observers, including IMF staff, have expressed concern that corrections to sustainable levels will likely require
large exchange rate adjustments, especially
against the U.S. dollar, with possibly disruptive
effects on global financial markets and economic activity.1 In contrast, other observers are
less concerned, arguing that a benign resolution
of global imbalances is likely with today’s deep
economic and financial integration.2
Some recent developments suggest that
globalization—the increasingly global dimension
of economic and financial transactions—has
changed the environment for external imbalances and their adjustment. For instance, it may
be argued that larger external current account
deficits or surpluses are the natural outcome of
the increased scope for cross-border trade in
financial assets, and that higher trade openness
and greater competition worldwide are likely to
have facilitated adjustment of global imbalances.
However, globalization has also brought new
challenges and risks. Larger external positions
raise economies’ exposure to financial market
disturbances, increasing the risks associated with
an abrupt realignment in investors’ expectations. Finally, the relationship between globalization and external adjustment can be more
ambiguous than suggested by casual observation.
Some aspects of globalization—including, for
example, more specialization in production—
may, at least in theory, hinder rather than facilitate adjustment.
Against this background, this chapter will
examine implications of globalization for external imbalances and their adjustment. The relationship between globalization and external
imbalances is complex and encompasses many
aspects. For tractability, the chapter will focus on
aspects that are particularly relevant from the
perspective of the unwinding of current imbalances, and the related risks. To this end, it is
organized in three parts.
• The first part discusses the rapid expansion of
two-way capital flows and the corresponding
increases in gross external asset and liability
positions and then examines the implications.
The chapter finds that these developments
have contributed to an environment in which
large current account surpluses or deficits can
emerge and be sustained and argues that this
can be helpful insofar as it allows for gradual
rebalancing. Moreover, while economies’
exposure to market and exchange rate
changes has increased with larger gross external positions, these valuation effects can, perhaps paradoxically, to some extent facilitate
external adjustment among industrial countries, as they are, in effect, wealth transfers
from countries with appreciating currencies to
countries with depreciating currencies. The
chapter notes, however, that these benefits
could turn into a liability if policies are not
consistent with a credible medium-term policy
framework aimed at external and internal balances, as expectations may not be well
anchored. In this case, investor preferences
Note: The main authors of this chapter are Thomas Helbling, Nicoletta Batini, and Roberto Cardarelli, with support
from Douglas Laxton, Dirk Muir, Panagiotis Konstantinou, Philip Lane, and Linda Tesar. Nathalie Carcenac provided
research assistance.
1See “How Worrisome Are External Imbalances?” Chapter II, World Economic Outlook, September 2002, and Obstfeld and
Rogoff (2001, 2004).
2See Cooper (2001, 2004).
109
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
may quickly change and the fallout from disruptive financial market turbulence would
likely be more elevated than it had been.
• In the second part, the chapter turns to real
globalization and examines how a broad fall in
trading costs has affected the magnitudes,
composition, and direction of trade flows as
well as other key determinants of external
adjustment. It finds that trade shares have
increased and the global distribution of trade
flows has become more equal, as emerging
markets have become more integrated, but it
notes that the empirical evidence on how real
globalization has affected price and demand
elasticities of trade flows is inconclusive.
• In the third part of the chapter, simulations
of the IMF’s new multicountry Global
Economic Model (GEM) are used to analyze
the combined effects of real and financial
globalization on the adjustment of external
imbalances. The chapter finds that real and
financial globalization should generally facilitate global rebalancing, provided financial
conditions remain benign, reflecting, among
other factors, the better burden sharing
implied by the more even distribution of
trade flows across the globe. It also shows,
however, that globalization has not fundamentally changed the nature of adjustment, nor
the magnitudes involved, and that larger net
foreign asset positions raise the potential risks
associated with unexpected changes in
investor preferences.
Overall, the chapter concludes that policymakers need to be very mindful of the risks associated with global imbalances, while at the same
time taking advantage of the scope that globalization provides to facilitate adjustment.
At the outset, two points should be noted.
• The chapter adopts a broad notion of globalization, its causes, and its impact. In particular,
while using as a starting point the narrow defi-
nition of an acceleration in the pace of growth
of international trade in goods, services, and
financial assets relative to the rate of growth in
domestic trade, the chapter also considers
phenomena such as the integration of emerging market economies, greater competition,
or reduced exchange rate pass-through,
which—while related to globalization—also
reflect other factors, including more credible
monetary policy frameworks.
• The chapter is not intended to cover the specific policies needed for the orderly resolution
of current imbalances, which are discussed in
Chapter I, or how the current imbalances have
emerged (see Chapter II, World Economic
Outlook, September 2002, and Hunt and
Rebucci, 2003, for recent discussions).
Moreover, it focuses mostly on industrial and
key emerging market countries that are highly
integrated and likely to play a major role in
the rebalancing.
Financial Globalization
Financial globalization—the global integration of capital markets—has accelerated noticeably since the early 1990s, as illustrated, for
example, by the rapid simultaneous increase in
many countries’ foreign assets and liabilities
(Figure 3.1).3 The trend toward larger external
assets and liabilities has been particularly relevant for industrial countries, where, relative to
output, both average external assets and liabilities about tripled between 1990 and 2003, reaching levels of more than 200 percent by the end
of the period.4 While the broad trend for emerging market countries has been similar, average
increases for these countries have been smaller
since the mid-1990s and, on a global scale, their
gross external positions remain relatively small
compared with those of industrial countries
(Table 3.1).
3See Lane and Milesi-Ferretti (2003, 2005a). There are other measures of financial globalization, including price-based
measures or the correlation between saving and investment. See Obstfeld and Taylor (2004) for empirical evidence based
on other measures that suggest similar broad trends.
4The data on gross foreign assets and liabilities used in this chapter are taken from the latest version of the database
developed by Philip Lane and Gian Maria Milesi-Ferretti (and used in Lane and Milesi-Ferretti, 2005b).
110
FINANCIAL GLOBALIZATION
The recent bout of financial globalization is
partly associated with the decline in information
processing and dissemination costs that have
fostered cross-border trade in an expanding
variety of financial instruments through decreasing transaction costs (Figure 3.2). Domestic
and external financial liberalization have played
a major role since the early 1970s when the
current era of financial globalization began
after a long period of financial disintegration
(see Box 3.1 for a comparison of the current
era of globalization with earlier ones).5 Finally,
real and financial globalization tend to stimulate each other. Increased trade flows, for
example, tend to lead to larger gross capital
flows, reflecting trade finance, among other
factors.
This section will examine two issues related to
the surge in international financial transactions.
First, it will analyze whether investors’ increased
incentives for international portfolio diversification have reduced the extent to which financial
markets still restrict net international borrowing
(the financing need associated with current
account deficits) and net foreign liabilities (the
corresponding stock measure). Second, it will
investigate the extent to which larger holdings of
foreign assets and liabilities expose investors to
greater valuation risks and what this means for
external adjustment.
While data on external assets and liabilities
have greatly improved so that more systematic
empirical analysis is possible compared with
some 10 years ago, important caveats nevertheless remain (see Box 3.2 on data issues).
Globalization and Net Foreign Assets
Traditionally, investors place the bulk of their
financial wealth in domestic assets despite more
favorable risk-return profiles—before transaction
costs and taxes—of globally diversified portfolios
Figure 3.1. Financial Globalization Trends, 1970–2003
(Percent of GDP)
Global financial market integration has proceeded rapidly, especially in industrial
countries, since the early 1990s.
Average
250 Industrial Countries: Foreign
Assets
Standard deviation
Emerging Market Countries:
Foreign Assets
100
200
80
150
60
100
40
50
20
0
1970
80
90
2000
250 Industrial Countries: Foreign
Liabilities
1970
80
90
2000
Emerging Market Countries:
Foreign Liabilities
0
100
200
80
150
60
100
40
50
20
0
1970
80
90
2000
1970
80
90
2000
0
Sources: IMF, International Financial Statistics; IMF, Balance of Payments Statistics; Lane
and MiIesi-Ferretti (2005b); and IMF staff calculations.
5See, among others, Edey and Hviding (1995) or
Williamson and Mahar (1998). The end of the Bretton
Woods regime of fixed but adjustable exchange rates also
contributed, as trade in foreign exchange and related
instruments began to spiral with floating rates.
111
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Table 3.1. Gross Assets and Liabilities, 1980–2003
(Billions of U.S. dollars, excluding foreign assets held by central banks)
External Assets
Industrial countries
North America
United States
Europe
France
Germany
United Kingdom
Asia and Pacific
Japan
Emerging markets and other
developing countries
Emerging Asia
China
Korea
Malaysia
Thailand
Latin America
Argentina
Brazil
Mexico
Others
External Liabilities
Industrial countries
North America
United States
Europe
France
Germany
United Kingdom
Asia and Pacific
Japan
Emerging markets and other
developing countries
Emerging Asia
China
Korea
Malaysia
Thailand
Latin America
Argentina
Brazil
Mexico
Others
1980
1985
1990
1995
2000
2003
2,287
668
584
1,474
203
216
519
145
134
3,975
1,310
1,185
2,232
265
308
838
433
410
9,701
2,331
2,119
5,522
722
1,017
1,695
1,848
1,778
15,334
4,107
3,754
8,659
1,292
1,579
2,342
2,568
2,445
26,810
7,914
7,265
16,066
2,323
2,504
4,400
2,830
2,605
36,039
8,454
7,680
24,311
3,317
3,867
6,293
3,273
2,922
114
28
...
4
4
1
55
4
10
9
31
193
72
18
7
7
2
80
5
19
21
42
366
166
25
17
5
7
126
6
29
37
74
711
384
47
58
13
14
214
3
51
50
113
1,479
948
225
103
39
22
325
3
74
45
206
1,849
1,193
273
127
53
23
386
2
96
30
270
2,485
708
525
1,578
193
224
508
198
147
4,186
1,464
1,206
2,311
306
289
755
411
307
10,531
2,909
2,459
5,835
845
751
1,762
1,787
1,529
16,139
4,849
4,274
9,096
1,385
1,535
2,426
2,194
1,812
28,419
9,741
8,965
16,386
2,214
2,534
4,497
2,293
1,808
39,039
11,452
10,476
24,790
3,247
3,794
6,429
2,797
1,979
552
138
...
33
12
10
267
29
90
66
147
905
284
27
59
31
20
395
54
121
108
226
1,298
493
77
55
31
41
492
70
153
134
313
2,378
1,125
255
147
81
138
819
144
241
230
435
3,527
1,626
479
195
101
116
1,247
236
403
309
654
4,208
2,116
659
293
115
111
1,310
192
420
362
782
Sources: Lane and Milesi-Ferretti (2005b); IMF, International Financial Statistics, and IMF staff calculations.
(the so-called home bias in asset holdings).6
Clearly, with globalization, opportunities for
international diversification have improved, as
important obstacles, such as high cross-border
transaction and information costs or regulatory
barriers, have been reduced.7 One would therefore expect that the home bias has decreased at
the global level.
6Since asset returns are only partially correlated across countries, investors may reduce risks that are specific to their
home country with international diversification (e.g., Solnik, 1974, or Obstfeld and Rogoff, 1996).
7Explanations of the home bias focus on factors reducing the incentives for international diversification, including high
transaction costs in cross-border transactions compared with domestic transactions, problems of cross-border information
dissemination, differences in regulatory regimes and regulatory barriers (e.g., regulations restricting foreign investment by
pension funds), and differences in consumption baskets, owing to the presence of transport costs (nontraded goods), but
112
FINANCIAL GLOBALIZATION
The decline in home bias matters for external
imbalances and their adjustment because it
determines the extent to which desired current
account balances8—which depend on factors
such as productivity growth differentials or
demographic changes—are accommodated by
international financial markets.9 If home bias
is strong, global demand for foreign assets will
be low and price-inelastic. Large issuers of
foreign liabilities will thus face high yields;
this will discourage net external borrowing,
and actual current account balances will likely
be smaller than desired ones. On the other
hand, if home bias is small, demand for foreign
assets will be higher and more price-elastic,
and larger net external liabilities will be less
costly. That said, net external borrowing will
remain limited by solvency considerations:
countries need to be able to amortize external
liabilities.
Over the past two decades, there is clear evidence that the home bias has declined and that
restrictions on net external borrowing have
eased.10
• Portfolio holdings of foreign bonds and
equity in some major industrial countries,
such as Canada, Germany, Japan, and the
United Kingdom, have clearly increased compared with domestic market capitalization
(Table 3.2).
• External current account deficits or surpluses
(relative to domestic incomes) have, on
also differences in taste. See, among others, French and
Poterba (1991), Tesar (1993), Tesar and Werner (1995),
Baxter and Jermann (1997), Baxter, Jermann, and King
(1998), or Obstfeld and Rogoff (2001).
8Defined as current account balances that would prevail
with no restrictions on international capital market access
and an infinitely elastic supply of capital.
9Financial globalization is a necessary condition for
larger current account deficits or surpluses but not necessarily a main cause. For example, international risk diversification alone may not generate net external borrowing
or lending: domestic investors can acquire foreign equity
with the proceeds from selling domestic equity to foreign
residents. Gross capital flows will increase, but inflows are
exactly matched by outflows.
10Evidence on the degree of remaining restrictions is
broadly similar for other measures. See also footnote 3.
Figure 3.2. Determinants of Financial Globalization
Decreasing communication and information costs and reduced restrictions on
capital flows have fostered financial globalization.
120 Communication Costs
(1990 U.S. dollars)
100
Satellite
charges
80
Internet Users
(percent of total population)
United
States
European
Union
50
40
60
30
Japan
Emerging
Asia
40
20
60
10
Telephone1
0
1930
50
20
World
70
90
1990
95
2000
0
Emerging Market Countries:
Capital Controls2
(percent of all countries)
Industrial Countries: Capital
Controls 2
(percent of all countries)
No restriction
Restriction
100
100
80
80
60
60
40
40
20
20
0
70s
80s
90s
00s
70s
80s
90s
00s
0
Sources: Busse (2003); IMF, Annual Report on Exchange Arrangements and Exchange
Restrictions (2004); World Bank, World Development Indicators; and IMF staff calculations.
1 Cost of a three-minute phone call from New York to London.
2 Restrictions on international financial transactions.
113
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Table 3.2. Overseas Portfolio Investment
(Percent of domestic market capitalization)
1970
1975
1980
1985
1990
1995
2000
2003
Canada
Portfolio investment
Equity
Bonds
2.0
3.1
0.7
1.9
3.2
0.6
2.1
3.6
0.4
2.4
3.5
1.3
6.0
9.6
1.9
12.9
25.4
2.2
18.7
29.3
3.2
14.3
21.2
3.6
Germany
Portfolio investment
Equity
Bonds
4.9
...
...
2.4
...
...
2.7
...
...
5.8
...
...
10.2
...
...
14.5
16.9
13
30.0
37.8
23.0
31.1
42.1
25.7
...
...
1.3
...
2.0
...
6.9
...
10.7
2.2
12.1
4.0
13.6
8.3
16.7
9.9
United Kingdom1
Portfolio investment
Equity
Bonds
9.5
...
...
8.6
...
...
11.4
...
...
27.5
...
...
34.0
33.1
35.6
37.1
33.5
43.4
42.6
40.9
46.4
48.1
52.4
42.1
United States
Portfolio investment
Equity
Bonds
1.5
0.8
2.7
2.1
1.1
3.1
2.3
1.3
3.3
2.2
2.0
2.4
3.5
5.6
2.1
6.4
9.3
3.5
7.8
10.5
3.8
7.4
12.7
2.3
Japan
Portfolio investment
Equity
Bonds
Sources: Tesar and Werner (1995); Lane and Milesi-Ferretti (2005b); IMF, Balance of Payments Statistics; national flow of funds and balance
sheet statistics; and IMF staff calculations.
Note: This table extends Table 2 in Tesar and Werner from 1990 to 2003 using similar but not necessarily identical data sources, replacing
1990 values with new data if available.
11991 values rather than 1990.
average, increased while their dispersion
across countries has widened in industrial
countries and, to a lesser extent, in emerging
market countries (Figure 3.3).
• Net external positions have, on average,
widened also, as has their dispersion on
account of the larger and more persistent
current account deficits and surpluses.
A simple way to quantify the decline in home
bias is to compare each country’s actual share of
foreign portfolio assets in its total portfolio asset
holdings with the share of other countries’
assets in the world total of assets (the world
market portfolio). If the former is smaller than
the latter, there is a home bias according to the
so-called international capital asset pricing
model (ICAPM).11 While some of the underlying assumptions are clearly unrealistic, the
model nevertheless provides useful benchmarks. The calculations shown in Table 3.3
suggest that between 1990 and 2003, the home
bias in bond and equity portfolio holdings of
most major industrial countries—except
Japan12—has declined but not disappeared.
This assessment is obviously tentative, given
that the underlying evidence is model-specific
and limited to patterns in major industrial
countries only. Nevertheless, the broad conclusion is similar to that reached in other recent
studies.13
11The ICAPM implies that investors should allocate their risky assets in proportion to the world market portfolio in equilibrium since other allocations involve idiosyncratic risks for which investors will not be compensated. See, among others,
Adler and Dumas (1983), Branson and Henderson (1984), and Harvey (1991).
12In Japan, the home bias has increased despite the increase in the actual share of foreign assets, as the benchmark share
implied by the ICAPM has risen even faster because of the relative decline of Japan’s share in the world market portfolio.
13See, among others, Obstfeld (2004), Lane and Milesi-Ferretti (2004a), and Engel and Matsumoto (2004). In contrast,
Heathcote and Perri (2004) argue that the home bias is much smaller than widely thought because their model implies a
lower optimal allocation of financial wealth in foreign assets compared with other models.
114
FINANCIAL GLOBALIZATION
Table 3.3. Portfolio Diversification: Actual
Foreign Shares and Benchmark Foreign Shares
Implied by Other Countries’ Share in World
Market Portfolio
(Percent)
1990
Equity
_________________________
1995
2000
2003
Bonds
___________
2000 2003
Canada
Actual
Benchmark
9.0
97.4
20.6
97.9
25.5
97.4
19.3
97.5
4.0
97.8
4.5
98.1
Germany
Actual
Benchmark
13.2
96.2
13.6
96.8
23.9
96.1
26.3
97.1
20.6
92.7
22.9
92.2
Japan
Actual
Benchmark
2.2
69.0
4.2
79.4
9.1
90.2
10.6
90.9
14.8
82.2
15.1
83.8
United Kingdom
Actual
29.5
Benchmark
91.0
30.1
92.1
38.4
92.0
45.7
92.0
62.0
95.9
69.4
95.4
United States
Actual
Benchmark
9.1
61.4
10.4
53.1
12.5
52.8
4.6
54.4
3.0
59.6
Figure 3.3. External Current Account Balances and Net
External Positions, 1970–2003
(Percent of GDP; absolute values)
5.7
67.5
Sources: Lane and Milesi-Ferretti (2005b); IMF, Balance of
Payments Statistics Yearbook and Global Financial Stability Report,
various issues; national balance sheet statistics; Standard and
Poors’ Emerging Markets Factbook, various issues; and IMF staff
calculations.
Note: The home bias can be gauged from the difference between
actual shares of foreign assets in total asset holdings and the
benchmark shares. Actual foreign shares are calculated as foreign
securities held as a share of total securities held by domestic
investors in each category. Benchmark foreign shares are based on
foreign countries’ share in total world market capitalization.
On average, external current accounts and net foreign assets have increased in
industrial and emerging market countries, suggesting that restrictions on net
external borrowing and lending have eased.
Average
5
Emerging Market Countries:
External Current Accounts
10
4
8
3
6
2
4
1
2
0
1970
50
If risk diversification is an important motive
for investors, one would expect that diversification across markets will be broad based.
According to the ICAPM discussed earlier,
investors should allocate their foreign assets
across countries according to their shares in
the world market portfolio. Table 3.4 compares
major industrial countries’ actual foreign
equity allocations across countries with ICAPM
benchmark allocations, taking the overall home
bias as given.14 The results suggest that diversification patterns are indeed broad based.
However, as the example of European countries
Industrial Countries: External
Current Accounts
Standard deviation
80
90
2000
Industrial Countries: Net
External Positions1
1970
80
90
2000
Emerging Market Countries:
Net External Positions1
0
50
40
40
30
30
20
20
10
10
0
1970
80
90
2000
1970
80
90
2000
0
Sources: IMF, International Financial Statistics; IMF, Balance of Payments Statistics; Lane
and MiIesi-Ferretti (2005b); and IMF staff calculations.
1 Net foreign assets as a percent of GDP.
14The results are based on data from the 1997 and
2002 Coordinated Portfolio Investment Survey conducted
under the auspices of the IMF. The general lack of long
time-series data prevents extensive historical analysis.
Geographical patterns in holdings of long-term bonds
are broadly similar.
115
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Table 3.4. Geography of Cross-Border Portfolio Equity Holdings
(Percent of total foreign portfolio equity holdings; second row contains benchmark shares implied by international capital asset
pricing model)
Source
____________________________________________________________________________________________
United
States
Canada
Japan
United
Kingdom
_____________
_____________
_____________
Euro area
Germany
_____________
Destination
1997
2002
1997
2002
1997
2002
2002
2002
1997
2002
United States
...
...
...
...
49.3
50.2
56.9
48.4
54.9
54.1
51.5
52.0
42.2
55.3
14.7
48.7
25.5
53.5
27.2
51.3
Canada
5.9
4.8
5.0
4.7
...
...
...
...
1.5
2.7
1.4
2.7
0.4
2.9
—
2.5
0.8
2.7
—
2.7
Japan
11.3
18.8
12.2
17.2
6.0
9.8
6.6
9.3
...
...
...
...
6.8
10.6
2.3
9.4
13.6
10.5
10.1
9.9
Euro area
31.2
24.9
26.8
27.5
13.5
13.0
13.3
14.9
16.6
14.1
16.0
16.0
...
...
64.0
11.9
35.4
13.9
32.5
15.8
Germany
5.4
7.0
2.9
5.6
...
3.7
...
3.0
5.5
3.9
2.3
3.2
...
...
...
...
6.4
3.9
4.4
3.2
United Kingdom
18.0
16.9
23.9
15.1
1.2
8.9
11.0
8.2
11.9
9.6
12.6
8.8
23.0
9.3
10.3
8.2
...
...
...
...
Industrial countries
79.2
77.3
78.5
75.6
87.5
88.1
93.2
86.8
91.2
87.2
87.3
85.9
92.7
85.0
97.9
86.6
87.7
87.3
80.9
86.0
Sources: French and Poterba (1991); IMF, Coordinated Portfolio Survey database; Standard and Poors’ Emerging Stock Markets Factbook,
various issues; and IMF staff calculations.
Note: U.S. investors’ holdings of Canadian equity amounted to 5.9 percent of their total foreign holdings in 1997. The benchmark allocation
predicted by the ICAPM for a given total foreign allocation would have been 4.8 percent.
shows, forces of “gravity” are also relevant.
Allocations to closely located countries or
regions tend to exceed benchmark allocations,
reflecting, among other factors, merchandise
trade patterns and the fact that shorter distances and cultural similarities appear to facilitate financial transactions, possibly through
their effects on transaction costs and information asymmetries.15
Despite growing U.S. net external liabilities,
European countries tend to hold less U.S.
equity than implied by the international capital
asset pricing model benchmark. More generally,
unlike in the 1980s, the share of U.S. portfolio
equity liabilities in total foreign portfolio equity
assets of other countries has been somewhat
below ICAPM benchmarks in recent years (see
also Bertaut and Kole, 2004). These observations suggest that more international financial
diversification has led to increased gross capital
inflows to the United States (in U.S. dollar
terms)—given greater overall flows at the global
level—but not to such an extent that investors
are now overweight in U.S. equity (or bonds).16
15The so-called gravity model of international trade suggests that, everything else being equal, neighboring countries
tend to have closer trade linkages than more distant countries (see Chapter II, World Economic Outlook, September 2002, for
a more detailed discussion). Recent research has found that trade-style gravity equations also perform well in explaining
bilateral investment patterns if augmented with financial market-specific variables (see Bertaut and Kole, 2004; Faruqee, Li,
and Yan, 2004; and Lane and Milesi-Ferretti, 2004a).
16This could suggest that the United States has become a relatively less attractive destination for foreign investors for at
least two reasons. First, with the weakening of business cycle linkages among major economic areas during the 1990s
return correlations declined compared with the 1980s. This, in turn, has made broad-based diversification more attractive
(see Heathcote and Perri, 2004). During the 1980s, diversifying internationally with U.S. assets only was perhaps more
attractive, given (1) deep and well-developed U.S. markets and (2) smaller benefits of more broad-based diversification
because of higher cross-country return correlations. Recently, however, return correlations have again increased. Second,
on the supply side, other securities markets have developed rapidly, as manifested in the growth of outstanding issues,
notably in the euro area. This has also increased the scope for broad-based diversification.
116
FINANCIAL GLOBALIZATION
Figure 3.4. Industrial Countries: Long-Term Real
Interest Rates and Net Foreign Assets1
In industrial countries, during the past decades, the relationship between real
interest rates and net foreign assets has weakened, as investors appear more willing
to hold foreign assets.
Average 1982–92
Average 1993–2003
10
8
6
4
Interest rates (percent)
Naturally, these observations are based on the
behavior of portfolio investment flows only, and
general conclusions about the future willingness of investors to hold U.S. assets cannot
be drawn.
Analyzing the long-run relationship between
real interest rates on long-term government
debt—a benchmark for rates of return in a
country—and overall net external positions provides a useful complementary perspective. With
home bias in asset demand, investors should
only be willing to increase the share of foreign
assets in their wealth if they are compensated
with increasing returns. Accordingly, real interest rates in countries with net external liabilities
should, on average, be higher than in countries
with net external assets, suggesting a negative
correlation between real interest rates and net
foreign assets.17 As noted earlier, if the home
bias broadly declines, such portfolio balance
effects should weaken, and the negative correlation between real interest rates and net foreign
positions should decrease in absolute terms. The
evidence shown in Figure 3.4 suggests that this
has indeed happened. The correlation between
the two variables during 1993–2002, while still
negative, was clearly smaller than during
1982–92.
Overall, financial globalization has created an
environment where net external borrowing and
lending are less restricted and where maintaining larger net foreign liabilities appears to
involve relatively lower costs. This can be helpful when it comes to external adjustment and
global rebalancing. For example, everything else
being equal, the United States now appears
more likely to be able to sustain larger net foreign liabilities in the long run at a lower cost
than, say, some 20 years ago. This could allow
for a more gradual adjustment of the same
2
-3
-2
-1
0
1
2
3
4
0
Net foreign assets to exports ratio
Sources: IMF, International Financial Statistics; Lane and MiIesi-Ferretti (2005b); and
IMF staff calculations.
1 Real interest rate calculated as nominal long-term interest rate at end of year t
minus the actual inflation rate at end of year t + 1.
17See Lane and Milesi-Ferretti (2002b) for a recent
empirical study presenting similar materials. Real
exchange rate changes are another source of return differentials (in the home currency of the investor). See
Branson and Henderson (1984) on the portfolio balance
approach and the role of home bias therein.
117
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Box 3.1. External Imbalances Then and Now
The current environment of globalized financial markets, which began in the 1970s for the
advanced countries and in the 1980s for the
emerging market countries, had an important
precedent in the four decades before 1914—the
era of the classical gold standard. Both eras of
financial globalization share common features,
including large net capital flows, but there are
also important differences, including today’s
much larger two-way capital flows.1 This box
compares salient features of financial globalization during 1870–1914 with those of today,
examines whether the large net capital flows in
the earlier era represented a global imbalance
comparable to that of today, and seeks to establish lessons from the earlier era.
Important common features in both eras of
globalization include the following.
• Large net capital flows and current account deficits
and surpluses. The 50 years before World War I
saw massive net private flows of capital from the
core countries of Western Europe to recent settlements overseas (mainly the rapidly developing Americas and Australasia). At the peak, the
associated current account surpluses in Britain
reached 9 percent of GDP and were almost as
big in France, Germany, and the Netherlands
(see the figure). For the principal capital
importers in the late nineteenth century
(Argentina, Australia, and Canada),2 current
account deficits exceeded 5 percent of GDP on
average. By comparison, over the past two
decades, current account surpluses and deficits
have been, on average, increasing, as discussed
in the main text, but they are still smaller than
during the gold standard era, in industrial and
emerging market countries alike. Another striking feature of the pre-1914 data is the high persistence in current account imbalances, even
when compared with today’s relatively persistNote: The main author of this box is Michael Bordo.
1See Bordo, Eichengreen, and Kim (1998), and
Obstfeld and Taylor (2004).
2Earlier in the century, the United States experienced similar net inflows but by the end of the nineteenth century, the country began running current
account surpluses.
118
Current Account Balances
(Percent of GDP)
Surplus Countries
(during gold standard)
10
Germany
5
0
France
-5
United
Kingdom
-10
1870 80 90 1900 10 20 30 40 50 60 70 80 90 2000
Deficit Countries
(during gold standard)
-15
15
Canada
10
5
0
-5
-10
-15
Australia
-20
-25
Argentina
1870 80 90 1900 10 20 30 40 50 60 70 80 90 2000
-30
-35
Sources: Obstfeld and Taylor (2004); and IMF staff calculations.
ent ones (Bordo, Eichengreen, and Kim, 1998;
and Obstfeld and Taylor, 2004).
• Current account reversals. Although current
account imbalances were generally more longlived in the pre-1914 era than in the recent
period, they were punctuated in some countries
by severe reversals, especially in the crisis-ridden
1890s (Bordo and Eichengreen, 1999). Current
account reversals have reemerged in today’s era
of financial globalization. In fact, Bordo and
others (2001) argue that the total incidence of
financial crises has been greater during the
post–Bretton Woods period than during the earlier period, although the output losses from
crises were somewhat larger pre-1914.
There are also important differences between
the two eras.
FINANCIAL GLOBALIZATION
• Distribution of current account deficits and surpluses.
Under the gold standard, countries of new settlement—the emerging markets of the time—ran
current account deficits while the major European economies had surpluses. In the current
era, core industrial countries run either persistent deficits or surpluses, with domestic savinginvestment imbalances redistributed primarily
among industrial countries rather than from
the core to the periphery as in the earlier era.
• Gross external positions are generally larger today.
Gross external positions were very close to net
positions before 1914—that is, net creditors
had large foreign asset positions whereas net
debtors had large liabilities. In contrast, most
major industrial countries today are both major
creditors and debtors irrespective of their net
position. The earlier pattern reflects the prevalence of long-term investment by the core countries in the countries of new settlement, seeking
higher returns by financing railroads and other
infrastructure as well as budget deficits (especially in the form of bonds but also in the form
of foreign direct investment). The substantial
growth of two-way flows between advanced
countries since 1980 has been associated with
both international financial diversification and
intertemporal consumption smoothing, as discussed in the main text.
• The adjustment mechanism is different. The historical record shows that adjustment to the significant and persistent external imbalances in the
pre-1914 era occurred largely through the
Humean price-specie-flow mechanism of the
classical gold standard (Bordo, 1984). Gold
flows ensured that equilibrium was restored
through changes in money supplies, the terms
of trade, and real exchange rates. In contrast,
the global economy is now on a managed
floating exchange rate regime, and external
adjustment depends no longer on gold flows
but on changes in exchange rates and international reserves, along with relative price movements, short-term capital flows, and valuation
effects (see Obstfeld, 2004).
Despite the fact that external imbalances were
often larger and more persistent before 1914
than they are today, contemporaries in the earlier
era did not view this as a problem for two broad
reasons. First, they strongly believed that except
in extreme situations (e.g., wars) the adjustment
mechanism described above would always be
stabilizing. Second, the nature of foreign investment was quite different. Most of the long-term
flows were to countries with abundant natural
resources and land on the one hand, and scarce
labor and capital on the other. Returns on labor
and capital were thus higher than in the more
developed countries, with excellent prospects of
sustained rapid long-term growth. The activities
financed tended to be those in which information asymmetries could be most easily overcome:
railroads and government (Bordo, Eichengreen,
and Irwin, 1999). Many recipient countries
tended to have sound institutions and sound fiscal fundamentals, further reducing the likelihood
of default, and many adhered to the gold standard, which served as a signal of fiscal rectitude
(“a Good Housekeeping seal of approval” (Bordo
and Rockoff, 1996)). In addition, many recipient
countries were part of the British Empire, with a
de facto British government guarantee that virtually eliminated country risk. That said, not all of
the recipients of foreign capital had such sound
fundamentals. Many of the countries of peripheral Europe and Latin America were prone to fiscal and monetary instability. Their record of
defaults and currency crises often attenuated the
capital flows.
The large gross external asset positions among
today’s advanced countries with floating
exchange rates have little precedence in the
past, which suggests that exposure to market and
exchange rate risks during external adjustment
may be quite different. Nevertheless, with large
imbalances in both eras of financial globalization, the earlier era may still provide relevant lessons. Most prominently, the generally
remarkably smooth adjustment among the countries adhering to a stable and credible nominal
anchor—the gold standard—underscores the
important role of well-functioning and credible
nominal anchors and sound financial policies in
facilitating external adjustment.
119
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Box. 3.2. Measuring a Country’s Net External Position
The net external position of a country is the
difference between the country’s external
assets—the claims of a country’s residents on
nonresidents—and its external liabilities—the
claims of nonresidents on residents. These
claims are divided in broad categories, which
correspond to those in balance of payments
statistics: foreign direct investment; portfolio
equity securities; portfolio debt securities;
other assets and liabilities (such as bank loans,
trade credits, and currency deposits); and
financial derivatives.1 A country’s claims on
nonresidents also include the reserve holdings
of the central bank, which are classified separately. Gross external debt is given by the sum
of portfolio debt liabilities, debt liabilities in
the direct investment category, and other
liabilities.
Until a few years ago, data on external assets
and liabilities (the so-called International
Investment Position, or IIP) were reported by
most industrial countries and few emerging
markets. In recent years the number of reporting countries has increased exponentially and
now totals about 100 (even though coverage
for newly reporting countries is typically
limited to the most recent period). The data
used in this chapter, constructed by Lane and
Milesi-Ferretti (2001, 2005b), combine country
estimates of external assets and liabilities (as
reported in their IIP) with estimates from
alternative sources (such as the World Bank’s
debt database for external debt liabilities) or
based on cumulative capital flows with appropriate valuation adjustments. The data cover 87
countries, including virtually all advanced
Note: The main author of this box is Gian Maria
Milesi-Ferretti.
1The FDI category reflects a “lasting interest” of an
entity resident in one economy in an enterprise resident in another economy (IMF, 1993). This includes
greenfield investment as well as significant equity participation (typically set at above 10 percent), while
remaining holdings of equity securities are classified
under portfolio equity investment. This implies that in
certain cases the distinction between these two categories can be blurred.
120
and emerging economies, for the period
1970–2003.2
How do valuation adjustments work? For
example, in the absence of information on foreigners’ holdings of domestic equities, these
can be approximated by cumulative net foreign
purchases of domestic equity (which can be
obtained from the widely available balance of
payments data), adjusted each year for the
change in the value of existing liabilities due to
fluctuations in stock prices and exchange rates.
These fluctuations in value, which can be
approximated by the variation in a domestic
stock price index, are not reported as investment returns in balance of payments statistics,
which for equities only record the flow of
dividends.3
As external assets and liabilities grow, these
valuation changes become quantitatively very
important—indeed, as discussed in the text, the
relationship between the current account and
the dynamics of the net external position has
substantially weakened in recent years. For
example, the U.S. net external position in 2003
was broadly unchanged as a ratio of GDP, since
the large current account deficit was offset by a
correspondingly large valuation gain, generated
by increased dollar values of U.S. foreign assets
as the dollar depreciated.
Valuation changes for portfolio equity and
FDI can be particularly large. For example,
between end-1998 and end-1999 the stock of
2A significant earlier contribution is Sinn (1990),
who constructs estimates of external assets and liabilities for the period 1970–87 for an even larger sample
of countries.
3Estimating valuation adjustments for the foreign
assets of a country is a more complex endeavor. A
precise calculation would require information on
the geographical and currency distribution of the
country’s claims, which is available for only a few
countries and typically for recent years. In the
absence of such information, one can for example
assume that the geographical distribution of assets follows the country’s trade pattern, or, for stock or bond
markets, relative market capitalization in the rest of
the world.
FINANCIAL GLOBALIZATION
Finnish portfolio equity liabilities increased
from about US$80 billion to about US$220 billion, even though net purchases by foreigners
during 1999 amounted to only US$10 billion!
The underlying cause for this valuation change
was the boom in the price of Finnish stocks—
particularly Nokia—during 1999.
Particular difficulties are posed by the valuation of foreign direct investment. Most countries
report the book value of their direct investment
assets and liabilities, while others, such as
France, Sweden, and the United States report
estimates both at book and at market value. The
difference can be significant, especially when
corporate valuations change substantially.
Finally, coverage of derivatives’ contracts in
international statistics is still very spotty, thereby
limiting our knowledge on the extent of crosscountry hedging.
With data on the stocks of external assets and
liabilities and the underlying capital flows, it is
possible to calculate the rate of return that a
country earns on its external assets and pays out
on its liabilities. These returns can be calculated by adding the yields on external claims
(which are measured as investment income
flows in the current account) and the capital
gain (which can be approximated by the difference between the change in value of the claim
and the underlying capital flow). On average,
rates of return are larger and more volatile
than yields, reflecting the fact that a significant
component of equity returns takes the form of
capital gains, rather than dividends (see the
figure).
The availability of comprehensive data on
external positions has enabled researchers to
address a number of important issues in international macroeconomics, including the determinants of long-run net external positions (Lane
and Milesi-Ferretti, 2002b), the link between net
current account deficit, which could facilitate
adjustment in production structures in the
United States and other countries and, thereby,
reduce the size of the overall exchange rate
United States: Rates of Return and Yields
on Foreign Assets and Foreign Liabilities
(Percent 1983–2003)
Foreign Assets
30
25
Return
20
15
10
Yield
5
0
-5
1983
88
93
98
Foreign Liabilities
-10
2003
30
25
20
Return
15
10
5
Yield
0
-5
1983
88
93
98
-10
2003
Source: IMF staff calculations.
external positions and real exchange rates
(Lane and Milesi-Ferretti, 2004b), and changes
in the extent of international risk sharing (Imbs,
2004; Huizinga and Zhu, 2004). Also, the sum of
external assets and liabilities has been used as a
volume-based measure of international financial
integration when studying the effects of integration on macroeconomic performance (see, for
example, Edison and others, 2002; and Prasad
and others, 2003).
adjustment (Obstfeld and Rogoff, 2000). On the
other hand, this opportunity could also turn
into a liability if macroeconomic policies do not
remain consistent with a credible medium-term
121
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Box 3.3. Financial Globalization and the Conduct of Macroeconomic Policies
Note: The main author of this box is Irina Tytell.
122
Financial Globalization and
Macroeconomic Policies1
Log Inflation
6
4
2
0
0
100
200
300
400
Log inflation, percent a year
8
-2
500
Gross foreign assets and liabilities, percent of GDP
Budget Deficit
8
6
4
2
0
-2
-4
0
100
200
300
400
Budget deficit, percent of GDP
Macroeconomic policies around the world
have improved over the past two decades. The
average budget deficit across both industrial and
developing countries has declined from approximately 5 percent of GDP during the late 1970s
to just over 2 percent of GDP recently. Similarly,
with monetary policy increasingly focused on
inflation control, inflation rates have been
decreasing across the globe.
A widely quoted explanation for these developments is that financial globalization has
exerted a disciplinary effect on the conduct of
policies, because international capital flows
adversely respond to imprudent macroeconomic
policies (e.g., Fischer, 1997; or Stiglitz, 2000).
This explanation is not universally accepted,
however, and some have argued that global
financial markets fail to discipline policies (e.g.,
Rodrik, 2001). Against this background, this box
reexamines the foundations of the hypothesis of
the disciplinary effects of financial globalization
and assesses its empirical merits.
From a theoretical perspective, the incentives
for host governments to conduct good policies
depend on their rewards and costs. To the
extent that good policies attract capital flows
that help raise the domestic capital stock, the
associated higher output is the reward. The
costs of good policies to policymakers are
related to political economy considerations.
For example, the need to conduct prudent
fiscal policy can limit politicians’ scope for discretion related to their own narrow interest. If,
on balance, the rewards are large enough to
offset the costs, globalization will indeed be a
disciplinary device. This in turn suggests two
conclusions.
• The disciplinary effect of financial globalization may be stronger for some policies and
weaker for others. In particular, if prudent fiscal policy exerts higher political economy
costs than, say, monetary policy, one would
expect the disciplinary effects of globalization
to be stronger for monetary policy.
-6
500
Gross foreign assets and liabilities, percent of GDP
Sources: Tytell and Wei (2004); and IMF staff calculations.
1For the period 1990–99.
• A critical issue regarding the benefits of capital flows concerns the possibility of changes in
investor sentiments in international capital
markets that are reflected in capital flow fluctuations unrelated to policies or developments in the host country. Through their
potential to lower the benefits of good policies, they tend to weaken the disciplinary
effect on policy conduct.
What does the empirical evidence for the
recent era of globalization look like? A simple
inspection of the relationship between a measure of financial globalization—the ratio of gross
foreign assets and liabilities to GDP—and inflation rates and budget balances suggests the following (see the figure).
FINANCIAL GLOBALIZATION
• The relationship between inflation and the
extent of financial globalization is generally
negative.
• The relationship between budget deficits and
financial globalization is markedly weaker.
Such bivariate relationships can, of course, be
misleading, since they do not control for other
determinants of macroeconomic policies or
other dimensions of globalization. For a more
complete analysis, Tytell and Wei (2004) used
two different econometric approaches.1
First, inflation rates and budget balances
were simultaneously regressed on financial
globalization (ratio of foreign assets and liabilities to GDP) and a number of other relevant
variables, including indicators of exchange rate
flexibility, central bank independence, government fragility and polarization, and trade openness. To isolate the effect of globalization on
policies while mitigating problems of reverse
causality (and measurement errors), the
authors focused on the common component
of international capital flows to countries in
the same geographic region. The results confirm that the coefficient on financial openness
in the inflation equation is negative and statistically significant (although fairly small in
magnitude). The same coefficient in the equation for budget balances is statistically insignificant. These results are robust to various
alternative specifications, different measures of
financial openness, and alternative instrumental
variables.
Tytell and Wei also use thresholds to classify
policies as “good,” “moderate,” or “bad,” to control for the fact that small fluctuations in inflation rates or budget balances are unlikely to
1The study covers a sample of 62 industrial and
developing countries, which excludes major oil producers and very small countries.
policy framework aimed at external and internal
balance. In this regard, it is important to note
that globalization may not be as effective a disciplinary device for the conduct of macroeco-
reflect changes in policymakers’ attitudes.2 This
allows an investigation of whether the disciplinary
effects differ depending on the economic situation and whether they work by inducing policy
shifts between different states rather than affecting policies within a given state. The results lend
support to the view that financial globalization has
a positive and statistically significant effect on the
probability of inflation decreasing from “bad” to
“moderate” and from “moderate” to “good” and
a negative effect on the probability of inflation
increasing from “good” to “moderate.” In contrast, the results do not provide support for the
view that financial globalization exerts a disciplinary effect on government budgets.
Overall, while plausible from a theoretical
perspective, the empirical evidence for the
hypothesis that financial globalization exerts a
disciplinary effect on the conduct of macroeconomic policies is rather mixed. There does
appear to be a significant, albeit small, impact
on monetary policy, but little on fiscal policy.
This suggests that relying on financial globalization to act as a device for ensuring policy discipline is not enough. It also suggests that other
factors must have played a role in the observed
improvement of macroeconomic policies, some
of which may have been related to globalization,
although through channels other than the one
discussed here. In this light, Box 3.4 examines
the link between globalization and monetary
policy from a broader perspective.
2The thresholds are based on the relevant literature.
For example, there is some agreement in the growth
literature that inflation rates beyond 7 to 11 percent
hurt growth in developing countries (e.g., Khan and
Senhadji, 2000). Hence, inflation rates of 10 percent
or less are classified as reflecting “good” monetary policy while inflation rates above 40 percent—a widely
accepted benchmark for high inflation—are classified
as “bad” policies.
nomic policies as is widely thought (see Box 3.3
for recent empirical evidence on the disciplinary effects of financial globalization on macroeconomic policies).
123
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Valuation Effects and External Adjustment
Figure 3.5. Valuation Effects on Net Foreign Assets
(Percent of GDP)
For larger industrial countries, annual valuation changes have become greater in
the 1990s, even if the cumulative sums shown below tend to be small due to the
frequent change in sign. For emerging market countries, magnitudes of valuation
effects have generally increased over the past decade.
1982–92
1993–2003
Selected Industrial Countries: Median of Annual Valuation Changes
(absolute values)
14
12
10
8
6
4
2
e
n
a
a
s
n
n
ly
y
d
m
ds
do trali Franc anad Spai Japa State rman Ita
ede land lan lan
C
Sw Fin ether itzer King Aus
e
d
G
e
t
w
i
d
N
S
ite
Un
Un
Selected Industrial Countries
(absolute values of cumulative change)
0
70
60
50
40
30
20
10
0
d
ds and
lan lan
l
Fin ther itzer
e
N
Sw
lia Italy pan
en om ce
ny
ain
tes da
Sp Sta Cana rma wed ingd Fran ustra
Ja
e
S
d
K
A
e
G
t
i
d
ite
Un
Un
Selected Emerging Market Countries
(absolute values of cumulative change)
40
30
20
10
a
zil
bia ay Chile guay
t i n ico
en Mex Bra lom rugu
ra
U
Arg
Pa
Co
a
s sia rea
nd
ne
esi
on aila ppi alay Ko
M
I n d T h Phili
0
Sources: IMF, Balance of Payments Statistics; Lane and MiIesi-Ferretti (2005b); and IMF
staff calculations.
124
A critical, and often underappreciated, implication of recent financial globalization is that
with both foreign assets and foreign liabilities rising sharply, investors—and countries—are much
more exposed to capital gains and losses owing
to exchange rate and other asset price changes.18
For example, if all foreign assets are denominated in foreign currency, a 10 percent exchange
rate depreciation increases their domestic currency value by 5 percentage points of GDP if the
stock of assets is 50 percent of GDP and by 10
percentage points if it is 100 percent. Higher
gross positions also tend to generate larger valuation changes in net foreign assets, although
structure and relative size of assets and liabilities
also matter.
IMF staff estimates of valuation changes in
industrial and key emerging market countries
confirm that as a percent of GDP, annual and
medium-term19 valuation changes in net foreign
assets generally increased in magnitude during
the 1990s compared with the 1980s (Figure 3.5).
The increase in medium-term valuation changes
is particularly noteworthy in some smaller and
relatively more open industrial countries that
have seen rapid changes in gross or net positions
(e.g., Finland, the Netherlands, and Switzerland)
and in emerging market countries, particularly
those that have faced large, one-off changes in
exchange rates (e.g., Asian emerging market
countries during the 1997–98 financial crises).
In other countries, however, magnitudes have
decreased, including in those that experienced
larger exchange rate changes in the 1980s compared with the 1990s.
An important implication of persistent valuation changes is that external current account
surpluses or deficits could become relatively less
important determinants of net foreign asset positions (Obstfeld, 2004), as illustrated by the expe-
18See Lane and Milesi-Ferretti (2001, 2005a), Tille
(2003, 2004), Gourinchas and Rey (2004), and Obstfeld
(2004).
19Defined as the cumulative sum over a 10-year period.
FINANCIAL GLOBALIZATION
rience of a number of industrial and emerging
market countries during the 1990s (Figure 3.6).
A key question, of course, is whether valuation
effects can help in external adjustment. Given
the prominent role of exchange rates in this
regard, the focus will be on the exchange-raterelated valuation effects. (Historically, valuation
effects arising from other asset price changes
have also been large in some cases, including
during times of sharply reduced asset return correlations.) Two factors are important in this
regard. First, the nature of the exchange rate
changes matters. If they are unexpected, valuation effects on net foreign assets are lasting,
while those associated with anticipated changes
are not, because the latter tend to be reflected
in asset yields, which, in turn, offset the valuation effects through their impact on current
accounts.20
Second, the general structure of external
assets and liabilities—especially their currency
composition but also the nature of the underlying financial instruments—also plays an important role in determining whether valuation
effects contribute to external adjustment.
• In industrial countries, where foreign assets
tend to be denominated in foreign currency
and liabilities in domestic currency, valuation
effects arising from unexpected exchange rate
changes—including those related to portfolio
preference disturbances—tend to facilitate
adjustment because they can provide for burden sharing among countries. In countries
with currency depreciation, the domestic currency value of foreign assets increases and—
with the value of liabilities unchanged—net
foreign assets improve; the reverse takes place
in countries with appreciating currencies. If
Figure 3.6. Valuation Effects, Current Accounts, and Net
Foreign Assets
In the 1990s, the contribution of valuation effects to changes in net foreign assets
was large relative to current account balances, especially in small open industrial
economies and east Asian countries. More generally, the correlation between
changes in net foreign assets and current account balances weakened.
Selected Industrial Countries: Current Account and Valuation Effects
Contribution to Changes in Net Foreign Assets (NFA), 1993–2003
(percent of GDP; cumulative changes over period indicated)
United States
Current accounts
Japan
Valuation effects
Canada
United Kingdom
Germany
France
Italy
Netherlands
Spain
Ireland
Finland
Norway
Sweden
Switzerland
Australia
-80
-60
-40
-20
0
20
40
60
80
120
Selected Emerging Market Countries: Current Account and Valuation
Effects Contribution to Changes in NFA, 1993–2003
(percent of GDP; cumulative changes over period indicated)
Current accounts
Argentina
Brazil
Valuation effects
Mexico
Chile
Colombia
Uruguay
Paraguay
Korea
Indonesia
Philippines
Malaysia
Thailand
China
India
Turkey
-50
-40
-30
-20
-10
0
10
20
30
40
Rolling 10-Year Correlation Between Changes in NFA and Current
Accounts (averages by country groups)
Emerging Asia
Latin
America
50
1.0
0.8
0.6
0.4
Industrial
countries
20See Lane and Milesi-Ferretti (2005a) and Obstfeld
(2004). Consider, for example, the case of an anticipated
depreciation from the perspective of a foreign bond
investor. Assuming unchanged risk preferences, the lower
value of the principal (in foreign currency) due to the
depreciation requires higher yields to preserve the present value of the investment. With anticipated exchange
rate changes, return adjustments may be amplified by
portfolio reallocation.
100
0.2
0.0
1985
90
95
2000
-0.2
Sources: IMF, Balance of Payments Statistics; Lane and MiIesi-Ferretti (2005b); and IMF
staff calculations.
125
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Figure 3.7. Valuation Changes in Net Foreign Assets
and Real Effective Exchange Rates
Valuation effects on net foreign assets facilitated the external adjustment in the
mid-1980s (as indicated by the ellipses).
Valuation changes (percent of GDP; left scale)
Real effective exchange rate (+ = depreciation; 2000 = 100; right scale)
4
United States
2
180
4
160
2
Euro Area and Japan
160
140
0
140
0
120
-2
120
-2
100
-4
-6
4
1980 85
90
95 2000
Emerging Asia
100
80
-4
60
-6
120
4
110
2
80
1980
85
90
95 2000
Rest of the World
-2
-6
100
90
90
80
-2
80
60
1980 85
90
95 2000
120
0
70
-4
60
110
2
100
0
180
50
70
-4
-6
60
1980
85
90
95 2000
50
Sources: IMF, Balance of Payments Statistics; Lane and MiIesi-Ferretti (2005b); and IMF
staff calculations.
126
the direction of exchange rate changes
matches that of external adjustment—that is,
currency depreciation in countries where
external balances have to improve (deficit
countries) and appreciation in other countries—this is equivalent to a wealth transfer
from surplus to deficit countries, which, everything else being equal, can help reduce the
amount of trade adjustment needed (see
Appendix 3.1).21 This mechanism was evident
during the U.S. current account correction of
1985–88, when the U.S. dollar depreciated by
about 30 percent in real effective terms
(Figure 3.7).22 More recently, the valuation
adjustments associated with the U.S dollar
depreciation during 2002–03 have offset about
three-fourths of the cumulative U.S. current
account deficit over the same period.
• In emerging market countries, where some
foreign liabilities, especially debt liabilities,
tend to be denominated in foreign currency,
valuation effects from unexpected exchange
rate depreciations are likely to complicate
external adjustment, given the increase in the
domestic currency value of these liabilities. On
the other hand, currency appreciation tends
to have positive valuation effects and improve
net external positions, as illustrated in Figure
3.8 for selected Latin American and East Asian
countries.
For a systematic assessment of the role of the
valuation channel in external adjustment, IMF
21It is, of course, possible to hedge at least partly against
valuation effects by using financial derivatives, which may
reduce the real impact of valuation effects, including
those arising from the wealth transfer from surplus to
deficit countries. This depends importantly on whether
the hedging is undertaken within countries or across
countries. In the case of the former, there are no aggregate hedging gains or losses at the country level because
the gains from hedging for some residents are offset by
hedging losses of other residents, and valuation effects
have real aggregate effects. Cross-border hedging activities, however, affect the real impact of valuation changes.
22If the currency realignment of the mid-1980s had
involved today’s larger gross positions, U.S. valuation
gains would have been more than twice as large as those
in the 1980s, while the losses in Japan and the euro area
countries would have been three times as large.
FINANCIAL GLOBALIZATION
staff undertook an econometric investigation of
the recent experience of 49 countries—21 industrial and 28 emerging market and developing
countries—for the period 1970–2003 (see
Appendix 3.2 for details and documentation of
the results). Following Gourinchas and Rey
(2004) and Corsetti and Kostantinou (2004), the
approach was to examine the dynamic responses
of trade balances (defined as the sum of net
exports of goods and services) and net foreign
assets to external imbalances. The intuition
behind this approach is as follows. Because countries must ultimately be able to pay their debts—
the so-called intertemporal external budget
constraint—trade balances and net foreign assets
are tied together in the long run. For example, a
debtor country will need to have a trade surplus
in the long run that is large enough to cover the
cost of its net external liabilities.23 While trade
flows and net foreign assets can deviate from this
long-run relationship in the short term, over
time that relationship has to be restored through
an adjustment in net exports, in net foreign
assets (other than the changes arising from the
adjustment in net exports), or a combination of
both. If a significant part of the adjustment
occurs through changes in net foreign assets, this
is interpreted as evidence for the valuation channel contributing to external adjustment. The
results can be summarized as follows.
• For the majority of emerging market countries, adjustment came entirely through trade
(net foreign assets were not found to respond
to deviations from the long-run relationship
between the trade balance and net foreign
assets). This finding does not mean that valuation effects are unimportant in these countries. On the contrary, the results imply, for
example, that negative valuation shocks, such
as those from a depreciation of the exchange
rate, tend to permanently worsen net foreign
asset positions in these countries and the
Figure 3.8. Selected Emerging Market Countries:
Valuation Changes and Real Effective Exchange Rates
Exchange rate depreciations tend to have adverse valuation effects in emerging
market countries.
Valuation changes (percent of GDP; left scale)
Real effective exchange rate (+ = depreciation; 1990 = 100; right scale)
40
Argentina
200
30
200
150
150
0
10
0
100
-10
-5
100
-10
50
-20
50
-15
-30
20
Brazil
5
20
-40
10
1979
89
99
Mexico
0
-20
160
10
1979
89
99
Korea
0
160
8
10
120
6
120
4
2
80
0
80
-10
40
-20
0
-8
120
15
0
-2
40
-4
-6
15
1979
89
99
Philippines
10
100
5
80
0
60
-5
40
-10
-15
20
-20
0
1980
90
2000
1980
90
2000
Thailand
0
140
10
120
5
100
0
-5
80
-10
60
-15
40
-20
20
-25
-30
1980
90
2000
0
Sources: IMF, Balance of Payments Statistics; Lane and MiIesi-Ferretti (2005b); and IMF
staff calculations.
23Strictly speaking, the intertemporal budget constraint
implies that the net present value of future trade surpluses (including current transfers) must be equal to current net external liabilities (see Appendix 3.1).
127
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Figure 3.9. Determinants of Real Globalization
Declining trading costs and trade liberalization have driven real globalization.
Transport Costs
(1990 U.S. dollars)
120
100
80
Sea freight1
60
40
Air transport 2
20
1930
40
50
60
70
80
90
2000
Trends in Average Tariff Rates
(percent; unweighted)
0
35
30
Developing countries
25
20
15
Industrial countries
10
5
1980
82
84
86
88
90
92
94
96
98
Number of Regional Trade Agreements
0
80
60
40
20
0
1960–64 65–69
70–74
75–79
80–84
85–89
90–94
95–99 2000–04
May
Sources: Busse (2003); World Trade Organization; and IMF staff calculations.
adjustment has to come from an improvement
of net exports.
• In contrast, in a number of industrial countries, particularly the United States and Japan,
valuation changes do appear to play a role,
with changes in net foreign assets reacting systematically to deviations from the long-run
relationship with net exports.
For major industrial countries, using quarterly
data, the analysis was refined further by examining the extent to which deviations of trade balances and net foreign assets from their long-run
relationship can predict other variables involved
in external adjustment, especially net export
growth and valuation changes. For all countries,
the results show that adjustment came primarily
through long-run changes in trade flows, but for
the United States the evidence also suggests that
valuation changes were helpful in the short
term.24
Overall, these results suggest that, for all countries, exchange rate changes primarily affect
medium- and long-term trade adjustment,
although in the short term they appear to generate helpful valuation changes in some industrial
countries, particularly the United States.25 This
may reflect the fact that short-run exchange rate
changes in industrial countries tend to be unexpected and, hence, can have large valuation
effects, whereas persistent exchange rate
changes over the medium and long run have a
larger anticipated component, with smaller valuation effects.
Naturally, these results are based on a period
during most of which gross foreign positions
were relatively small compared with exports or
imports. One could therefore argue that analysis
based on past data likely underestimates the role
of valuation effects today.26 Overall, the results
1 Average ocean freight and port charges per short ton of import and export cargo.
2 Average air transport revenue per passenger mile.
24For the United States, the findings match earlier ones
by Gourinchas and Rey (2004) based on another data set.
25These results are consistent with—but not necessarily
evidence for—the notion that the United States enjoys a
reserve currency premium.
26In addition, there is the problem of using standard
econometric methods in periods of rapid change, as
parameters may change.
128
REAL GLOBALIZATION
Table 3.5. Changes in Geographical Trade Patterns
(Extra-regional trade flows as a percent of world GDP; excluding intraregional trade)
Origin
_______________________________________________________________________________________________
Emerging Asia
Euro area
Japan
Rest of world
United States
______________
______________
______________
______________
______________
Destination
1984
2003
1984
2003
1984
2003
1984
2003
1984
2003
Exports
Emerging Asia
Euro area
Japan
Rest of world
United States
Total
...
0.18
0.28
0.27
0.47
1.20
...
0.65
0.45
0.41
0.93
2.44
0.16
...
0.06
2.19
0.38
2.79
0.40
...
0.11
2.96
0.57
4.04
0.34
0.12
...
0.42
0.49
1.37
0.59
0.16
...
0.24
0.32
1.31
0.49
2.37
0.16
...
1.40
4.42
0.89
2.65
0.21
...
1.85
5.60
0.23
0.31
0.19
1.03
...
1.75
0.37
0.31
0.14
1.18
...
2.00
Imports
Emerging Asia
Euro area
Japan
Rest of world
United States
Total
...
0.16
0.34
0.49
0.23
1.22
...
0.40
0.59
0.89
0.37
2.25
0.18
...
0.16
2.37
0.38
3.08
0.65
...
0.21
2.65
0.41
3.92
0.28
0.06
...
0.54
0.22
1.10
0.45
0.11
...
0.33
0.11
1.06
0.27
2.19
0.42
...
1.03
3.91
0.41
2.96
0.24
...
1.18
4.79
0.47
0.39
0.49
1.40
...
2.74
0.93
0.53
0.33
1.80
...
3.60
Sources: IMF, Direction of Trade Statistics, and International Finance Statistics; and IMF staff calculations.
provide some evidence for the importance of
the valuation channel facilitating short-term
external adjustment in industrial countries, but
also suggest that trade balance adjustment
remains key to restoring external balance in the
long term, so that trade balances and current
account balances continue to be important as
indicators of external balance.
Real Globalization
Besides financial markets, globalization has
also profoundly affected markets for goods and
services, with global trade27 as a percent of
GDP increasing from some 20 percent in the
early 1970s to about 55 percent in 2003. A
broad fall in costs of global trading—including
declines in transport costs, costs of information
gathering and sharing, and continued decreases
in government-imposed trade barriers such as
tariffs—has been the key driving force behind
real globalization (Figure 3.9). Historically, the
current era of real globalization began earlier
than financial globalization, as liberalization of
external trade regimes started in the 1950s
under the umbrella of the General Agreement
on Tariffs and Trade (GATT). This section will
document these changes and try to shed light on
their implications for external adjustment.
Globalization and Changes in Trade Patterns
With the fall in trading costs over the last few
decades, global trade patterns have changed
markedly. First, geographical patterns of trade
have changed (Table 3.5).28 Arguably, in recent
years, the most important feature in this respect
has been the growing importance of emerging
market economies in world trade, especially—
but not only—the fast growing economies of
emerging Asia, while trade shares of major
industrial countries have decreased in relative
terms. An important implication of these
changes is that the trade adjustment associated
with the resolution of global imbalances will be
shared differently. In addition, as globalization
has contributed to strong growth in many
emerging market economies, relative economic
27Defined
as the sum of exports and imports of goods and services.
changes in geographical trade patterns have partly emerged because of new regional trade agreements (e.g., the
North American Free Trade Agreement) or the dismantling of old agreements (e.g., the Council for Mutual Economic
Assistance, or COMECON).
28The
129
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Table 3.6. Trade Intensity
(Manufacturing trade as a fraction of manufacturing GDP)
United States
EU-15
Japan
Canada
Mexico
Canada and Mexico
1970
1980
1990
2000–011
0.32
0.38
0.66
1.36
...
...
0.51
0.50
0.75
1.79
...
...
0.79
0.70
0.66
2.51
1.87
2.24
1.06
0.89
0.88
2.93
2.57
2.75
Source: OECD.
1For Canada and EU-15 manufacturing GDP is only available
through 1999. Ratios in the last column use trade data through
2001 but divide by 1999 value added. EU-15 refers to the member
countries of the European Union prior to May 2004 when several
eastern European countries joined the European Union.
sizes of major regions have changed, which has
helped to spread the spillover of shocks originating in one region to other world regions—with
potentially beneficial effects on external
adjustment.
Magnitudes and composition of trade have
also changed materially, with the following main
observable changes.
• Two-way intra-industry trade. The bulk of trade
now takes place within, not across, industries,
as countries tend to specialize in varieties of
particular goods, at the level of both final and
intermediate goods, rather than in a particular
industry. This partly reflects the fact that gains
from lower trade barriers arise not only from
lower prices, but also from the availability of a
wider range of similar goods that better accommodate individual needs of consumers and
firms. As a result, trade flows have increased
rapidly relative to domestic production—
reflected in higher trade shares—and more
countries, especially industrial countries, now
tend to be both large exporters and large
importers within narrow industry categories.
• Fragmentation in the production of manufactured
goods. With declining trading costs lowering
the costs of producing in multiple locations,
firms have begun to divide the production
process into multiple steps at different locations to take advantage of location-specific
advantages in each step (e.g., low labor costs
in the production of labor-intensive parts).29
Because each step involves imports and reexports of parts and intermediate goods up to
the final assembly, manufacturing trade has
increased dramatically compared with manufacturing GDP (Table 3.6).30
The ascent of multinational firms has played
an important role in overall trade growth and in
the changes in the composition of trade.31
Foreign direct investment (FDI) inflows and outflows are good indicators of the increasingly
international nature of firms (Table 3.7).
Recently, the acquisition of existing firms in
emerging markets, notably in Latin America and
Asia, by firms headquartered in industrial countries, has been a prominent feature of FDI
developments—mirroring changes in geographical trade patterns discussed earlier.
These changes in magnitudes and composition of trade related to real globalization may
have affected external imbalances and their
adjustment through a number of channels.
• Trade shares, trade home biases, and spillovers.
With real globalization, trade shares (trade as
a percent of domestic production) have generally been increasing. In many, but not all
cases, this has been reflected in a declining
share of domestically produced inputs utilized
in production, a development sometimes
referred to as the reduction of the home bias
in production (as opposed to the home bias in
international financial markets discussed earlier).32 As a result, spillover effects from market to market are likely to have become larger,
especially for sectoral disturbances, so that any
29Many terms are used to describe this phenomenon, including vertical specialization, production sharing, vertical production networks, and outsourcing. Hanson, Mataloni, and Slaughter (2003) find that, in a cross-section of countries, the
location of vertical processing networks is sensitive to local labor costs.
30See Yi (2003) and Chen and Yi (2003).
31See, among others, Burstein, Kurz, and Tesar (2004).
32Similar considerations apply to domestic demand for final tradable goods, as the share of such goods produced domestically has generally also fallen (reduction in “consumption home bias”).
130
REAL GLOBALIZATION
Table 3.7. Foreign Direct Investment Flows
(Period averages, percent of total, unless otherwise noted)
Outflows by Area of Origin
__________________________________
High-income countries
United States
Europe
Japan
Oceania
Total
Developing and transition countries
Latin America
Africa
Asia (excluding Japan)
Oceania
Central/eastern Europe
Total
World (millions of U.S. dollars)
Inflows by Area of Destination
__________________________________
1970
1980
1990
1970
1980
1990
45.99
42.33
5.40
0.91
98.98
20.98
51.63
13.90
2.17
93.84
21.79
44.61
5.41
1.08
87.5
11.19
43.79
0.63
5.97
75.93
29.90
35.64
0.45
4.71
75.19
20.10
38.79
0.57
2.33
65.23
0.32
0.41
0.29
...
0.02
1.03
0.93
1.44
3.76
0.01
0.01
6.16
2.19
0.50
9.34
0.01
0.27
12.50
12.66
4.29
6.60
0.45
0.01
24.08
9.02
2.32
13.00
0.17
0.26
24.81
11.10
1.75
18.66
0.07
3.05
34.77
23,678
124,407
523,293
20,956
113,917
530,174
Sources: Barba Navaretti and Venables (2004); and IMF staff calculations.
adjustment will have a magnified effect on
trade flows. A contraction in demand for a
particular final good, for example, will trigger
a contraction in demand for imported intermediate inputs all along the production chain.
Thus, trade flows will in general appear more
elastic with respect to demand changes.
Regarding net exports, however, implications
are less clear-cut, partly because both exports
and imports are generally affected.33
• Price elasticity of trade flows. The ability to purchase similar if not identical goods from
domestic and foreign sources could raise the
substitutability between foreign and domestically produced traded goods, which would
tend to make bilateral trade flows more
responsive to changes in prices, including
those arising from exchange rate changes. For
a given adjustment in quantities, relative
prices will then have to change less. However,
integrated production lines may have made
some firms more dependent on particular foreign inputs that have no close substitutes, and
their demand for foreign inputs may have
become more price-inelastic. Empirically,
there is little evidence to date that the elasticity of substitution between foreign and
domestically produced traded goods—which
determines the price elasticity of trade flows—
has changed (e.g., Ruhl, 2003).
• Tradability and the share of nontradable goods in
production. With declining trading costs, one
would expect the tradability of goods and services to have increased and the share of nontradable goods in demand and production to
have decreased.34 This would, everything else
being equal, facilitate adjustment, as the
required changes in tradables relative to nontradables output would be less and could occur
with relatively smaller changes in the relative
price of tradables compared to nontradables
(the real exchange rate). Empirical evidence
based on sectoral input-output tables, however,
suggests that, unlike in many emerging market
countries, the tradables sector share output in
most industrial countries has actually fallen
slightly in recent years because of the rapid
expansion of service sectors.35
33Production-sharing links do not automatically imply tight links among national economies. If the contribution of local
value added to the total value of trade is small, changes in trade volume could be large with very little impact on the local
economy (see Burstein, Kurz, and Tesar, 2004).
34See, for example, Bayoumi and others (2004) or Bravo-Ortega and di Giovanni (2005) for recent studies.
35See Batini, Jackson, and Nickell (2002) on the methodology.
131
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Other Implications of Real Globalization
Figure 3.10. Degrees of Foreign Competition and
Effective Exchange Rates in G-7 Countries
For the G-7 countries, foreign competition seems to have increased over time.
Ratio of other G-7 members' export deflators (PPP-weighted) to
country GDP deflator (right scale)
Nominal effective exchange rate (+ = depreciation; left scale)
1.2
Canada
2.0
1.2 France
2.8
1.8
1.1
2.4
1.1
1.6
2.0
1.0
1.4
1.0
1.6
1.2
0.9
0.9
1.2
1.0
0.8
1975 80
1.8
85
90
95
2000
Germany
1.6
0.8
0.8
1.6
1.5 Italy
1.4
1.3
1.2
1.0
1.0
0.8
0.8
0.5
1.2
1.2
1.1
1.1
1.0
1.0
0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.6
1975 80
85
90
95 2000
7.5
6.0
1.4
4.5
1.2
3.0
1.0
0.8
1975 80
3.2
85
90
95
2000
Japan
2.8
2.4
1.5
1975 80
85
90
95 2000
United Kingdom
3.0
2.4
1.6
1.2
0.8
0.4
1975 80
85
90
95
2000
United States
1.8
1.5
1.5
1.0
1.2
1975 80
85
1.8
1.2
85
90
95
2.1
2.0
0.5
1975 80
90
95
2000
0.9
Sources: IMF, International Financial Statistics; and IMF staff calculations.
132
0.0
3.6
2.0
2.5
0.8
2000
0.6
There are a number of other factors partly
linked to greater trade integration that may also
affect the way external imbalances adjust. These
include the following.
• Greater international competition and lower markups. Lower trading costs in recent decades
have spurred product market competition by
fostering the international supply of final and
intermediate goods. At the same time, a wave
of merger and acquisition activity, even in services industries (e.g., retailing, insurance, banking, and telecommunications) has brought
international cost-cutting techniques and price
competition into sectors that used to be isolated from such global pressures. Measuring
competition is difficult, but the ratio of world
export prices (in local currency terms) to the
GDP deflator—a frequently used indicator for
foreign competition—has decreased, which is
consistent with the notion that foreign competition has intensified in G-7 countries even
when changes in effective exchange rates are
considered (Figure 3.10).36 Evidence from sectoral markups partly supports this view
(Martins and Scarpetta, 1999) even though
the changes over the past two decades have,
on average, been relatively small. The ongoing
compression of markups puts a cap on global
price levels and inflationary pressures from
disturbances. This generally facilitates the task
of delivering low and stable inflation for monetary policymakers (see Box 3.4).
• Greater flexibility and lower rigidities. Structural
reforms over the past two decades—partly
reflecting pressures from greater international
36World export prices in local currency terms depend
on the nominal effective exchange rate by construction
(see Bailliu and Fujii, 2004). This implies that the real
price of exports can fall exclusively owing to an appreciation of the domestic currency. In the cases of Canada,
France, Italy, and the United Kingdom, however, real
prices of exports seem to have fallen—signaling an
increase in foreign competition—even in the face of constant or depreciating nominal exchange rates. In Figure
3.10, world export prices for each country are approximated by export deflators of other G-7 countries.
THE EFFECTS OF GLOBALIZATION: AN INTEGRATED PERSPECTIVE
competition—have enhanced economic flexibility, including by lowering real and nominal
rigidities in product and labor markets, especially in industrial countries (see, for example,
Chapter III, World Economic Outlook, April
2004). While recent empirical evidence,37 at
both the macroeconomic and the firm level,
suggests that rigidities of both kinds are still
present in most economies today, preliminary
results obtained estimating aggregate pricing
equations indicate that in most G-7 countries,
nominal or real price rigidities or both have
decreased somewhat since 1970 (Appendix 3.3
details these results). Other factors, such as
the adoption of “just-in-time” technologies to
reduce durables inventories relative to sales,
have also contributed to enhancing flexibility.
As result, economies are now believed to be in
a better position to absorb and rebound from
shocks, which would, by implication, also facilitate the rebalancing of global imbalances.
• Reduced exchange rate pass-through and improved
monetary policy credibility. Empirical evidence
suggests that the short-run pass-through of
nominal exchange rate movements to domestic final goods prices has declined in many
countries in recent years. This in part reflects
more credible monetary policies and the
associated transition to a lower inflation environment (see Taylor, 2000; Choudhri and
Hakura, 2001; Devereux and Yetman, 2002;
Gagnon and Ihrig, 2004; Bailliu and Fujii,
2004), as well as changes in pricing practices
of firms engaged in trade—such as an
increase in the prevalence of local currency
pricing (see Devereux, Engel, and Storgaard,
2004). In contrast, the pass-through of
exchange rate changes to prices of imports
“at the dock” is higher, especially in the long
run (e.g., Campa and Goldberg, 2002). While
the extent and causes of the decline in the
exchange rate pass-through to domestic consumer prices remain subject to debate (see,
among others, Obstfeld, 2002), a lower pass37See
38See
through can have positive and negative consequences for external positions and their
adjustment. On the one hand, it implies that
economies are more insulated from external
shocks, allowing growth to remain relatively
stable in the face of high exchange rate
volatility even in very open economies. On
the other side, a lower pass-through may
reduce the expenditure-switching effects of
exchange rate changes, thereby complicating external adjustment. The fact that today
a large portion of trade is intrafirm and
intra-industry, and that the exchange rate
pass-through to the prices of imported intermediates does not, in general, appear to have
changed much over time, suggests however
that the overall responsiveness of trade to
changes in exchange rates may not have
shifted dramatically, even if the exchange rate
pass-through to final consumer goods prices
has declined over the years.
In sum, real globalization has affected external imbalances and adjustment through a number of channels, although the extent and, in
some cases, direction are yet to be established.
Therefore, the chapter now turns to model simulations for a fuller and more integrated analysis
of how real globalization has affected external
adjustment.
The Effects of Globalization: An
Integrated Perspective
To investigate the implications of globalization for the adjustment of external imbalances,
this section uses simulations of the IMF’s
GEM.38 GEM incorporates many trade linkages
with an explicit microeconomic foundation, and
is thus particularly well-suited to analyze the
impact of real globalization discussed earlier.
Moreover, while financial linkages in GEM are
still cursory, the model nevertheless can mimic
key aspects of financial globalization such as
the apparent increasing willingness of investors
Angeloni and others (2004).
Bayoumi (2004) and Laxton and Pesenti (2003).
133
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Box 3.4. Monetary Policy in a Globalized World
This box looks at the implications of globalization for monetary policy. The main conclusion is that although globalization has altered
the environment in which central banks operate
and made it more difficult to assess and predict
economic developments, the objectives and
instruments of monetary policy are no different
in a more integrated world.
Effects of Globalization on Monetary Policy
Perhaps the most dramatic implication of
financial integration and freer capital flows for
monetary policy is the fact that it has become
harder to simultaneously maintain fixed
exchange rates and conduct an independent
monetary policy dedicated to domestic objectives. If a country’s exchange rate is pegged to
the currency of another country, then its interest rates will have to follow closely those of the
country to which its currency is pegged: any positive (negative) deviation would, in fact, trigger
capital inflows (outflows) putting upward
(downward) pressure on the exchange rate parity. The sheer increase in cross-border financial
transactions in the past 20 years has made it
harder for central banks to counter such pressures, even when they hold significant foreign
exchange reserves. Of course, countries can
decide to limit capital flows by imposing capital
controls, but this implies that they have to forgo
the benefits of capital integration, which puts
them at a disadvantage relative to countries with
floating exchange rates and free capital flows in
tapping world saving. The conflicting nature of
these policy options—fixed exchange rates,
independent monetary policy, and free capital
mobility—also known as the “impossible trinity,”
has pushed many countries in recent years to
abandon exchange rate pegs altogether (see
Fischer, 2001).
Globalization has also had numerous other
important effects on monetary policy.
• Greater international exposure. By making
economies more open, globalization has made
Note: The main author of this box is Nicoletta
Batini.
134
economies more exposed to international
shocks, thereby raising the challenges to which
monetary policy must respond. For example,
the fact that countries now trade more with
each other in both final and intermediate
goods and that production has become often
geographically very fragmented implies that
even the smallest demand and supply disturbances in a country can have repercussions for
production and profitability in countries elsewhere. In addition, under globalization, financial markets are increasingly integrated with
those abroad. Disturbances and policy decisions in one country are reflected swiftly in
markets around the world.
• Changes to the transmission of monetary policy.
There are two main ways in which this happened. First, while central banks can still control short-term interest rates under floating
exchange rates with financial globalization
(Boivin and Giannoni, 2002, 2003), deeper
financial integration has made exchange rates
more reactive to changes in interest rate differentials.1 All things being equal, this means
that changes in monetary instruments have a
larger impact on exchange rates, reinforcing
the exchange rate channel of monetary transmission. Second, globalization has influenced
the way exchange rate changes affect aggregate demand, even though the overall direction of such influence on the exchange rate
channel is yet unclear. On the one hand,
trade globalization has boosted import and
export volumes in many countries, and so
changes in the exchange rate now affect a
greater portion of aggregate demand. On the
other hand, globalization may have contributed to weaken the link between exchange
rates and the relative price of imports and
exports (the third section of this chapter
describes this process in more detail).
1Average correlations between (quarterly) changes
in the U.S. dollar/euro bilateral exchange rate and
nominal short-term interest rate U.S.–EU12, for example, have been strengthening since 1970. For a more
general discussion, see, among others, Panigirtzoglou
(2004) and Barnea and Djivre (2004).
THE EFFECTS OF GLOBALIZATION: AN INTEGRATED PERSPECTIVE
• Compressed markups and wages. Globalization
has increased international competition, lessening markups and helping contain wage
pressures. This has helped central banks
attain their goals because it has lowered inflation expectations in many countries, keeping
inflation subdued and, at the same time,
allowing economies to operate at a higher
degree of resource utilization with a lesser
threat of rising inflation.
• A more complex world. Finally, as globalization
has changed the behavior of consumers and
firms and altered the nature and number of
economic linkages across countries, historical
data have become a less reliable yardstick for
interpreting the present. On the whole, globalization has thus made it harder to model economies and predict economic developments.
Central banks’ increasing interest in how to
operate under uncertainty, and the boom of
analysis in this area in recent years (see Swiss
National Bank, 2003) are testament to this fact.
How Should Monetary Policy Account for
Globalization?
Does globalization introduce new objectives
for monetary policy? Probably not. Although
globalization has, in fact, reduced the ability of
central banks to pursue inconsistent objectives,
as exemplified by the dilemmas now posed by
the “impossible trinity,” it is generally agreed
that central banks should continue using their
sole policy instrument to achieve the primary
goal of price stability. They should therefore
refrain from pursuing additional objectives—for
example, external balance or exchange rate stability—that may be perceived as playing a more
prominent role in a globalized world but that
cannot be directly controlled by monetary policy
(Brash, 2001; Stevens, 2004; Greenspan, 2004).
Likewise, globalization does not seem to call
for new instruments. As economies are more
open, central banks may be tempted to influence demand and prices by making more intensive use of the exchange rate, either by affecting
interest rate differentials or by directly buying
and selling foreign currency to affect its value.
Clearly, with globalized economies the exchange
rate has become a more pivotal indicator of
monetary conditions and inflationary pressures.
However, trying to control inflation and output
by manipulating the exchange rate can be dangerous. Batini and Nelson (2000), for example,
show that when uncovered interest parity holds,
trying to move the exchange rate reduces
exchange rate variability but actually increases
the variability of inflation. The difficulties experienced by the Reserve Bank of New Zealand
and the Bank of Canada with Monetary
Condition Indices2 (MCIs) are telling in this
respect (see Batini and Turnbull, 2002).
On the other hand, monetary policy still
needs to adjust to globalization in various ways.
First, central banks should continue to work on
refining their analytical toolkits to take into
proper account developments from deeper
financial and trade integration and the economic implications thereof. As discussed above
and documented by Rogoff (2003), overall globalization seems to have helped reduce inflationary pressures in many countries, but the precise
mechanism and the magnitude of such effect
are still far from clear-cut, so more research is
needed on this front. Second, when doing policy analysis and economic evaluation, central
banks need to continue their efforts to move
from models in which parameters depend heavily on historical estimates toward models in
which parameters are less likely to change over
time (see Sargent, 1999; and Pagan, 2003).
Finally, central banks need to persevere in their
efforts to exchange information on economic
developments and discuss global economic
issues within the many international forums that
exist today. Given the growing interdependence
of national economies and macroeconomic policies, the coordination of such policies has
undoubtedly become more important.
2MCIs are fixed-weight weighted averages of interest
rates and exchange rates. Their use as operating targets implies moving one of the rates or both to bring
the average in line with some predetermined equilibrium that is assumed to be consistent with price stability and stable economic growth in the long run.
135
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Table 3.8. Differences in Parameters in Vintage Calibrations, 1980s and 2000s
Emerging Asia
_______________
Japan/Euro Area
_______________
Rest of World
_______________
United States
______________
1980s
2000s
1980s
2000s
1980s
1980s
2000s
5
7
5
4
66
74
65
87
30
33
10
10
94
95
96
97
57
61
33
30
35
42
32
32
15.4
28.6
14.3
26.7
23.5
41.7
21.3
40.0
19.0
34.5
17.5
32.9
16.7
31.3
15.0
27.9
Nominal rigidities (quarters)
Average price contract lengths
4
4
5
4
5
4
5
4
Monetary policy reaction function
Feedback parameters:
Lagged interest rate
Inflation gap
Output gap level
Real growth
1.00
—
—
—
1.00
—
—
—
0.75
0.44
0.19
0.14
0.81
0.52
0.10
0.51
0.75
0.44
0.19
0.14
0.81
0.52
0.10
0.51
0.75
0.44
1.19
0.14
0.81
0.52
0.10
0.51
Description
Home biases (percent)1
Consumption goods
Investment goods
Tradable sector output (percent of total output)
Markups (percent)
Tradables sector
Nontradables sector
2000s
Source: IMF staff calculations.
1Weights of domestically produced tradable goods in the total input of tradable goods in the production of consumption or investment goods.
to hold foreign assets, and the connection
between growing net foreign asset positions and
risk premia.39
To conduct the simulations, a new four-bloc
version of GEM was developed. The four-bloc
partitioning, which involves (1) the United
States; (2) the euro area plus Japan; (3) emerging Asia; and (4) the rest of the world, mirrors
the geographical constellation of the current
imbalances. In the simulation experiments, two
versions of the model are used. One version
replicates an environment of “low economic
integration” based on data from the 1980s, while
the second version portrays a “high economic
integration” environment based on 2000 data.
Key differences in the two model versions, resulting from or connected to globalization are
summarized in Table 3.8 and Figure 3.11.40
Compared with the “low economic integration”
version, the “high economic integration” version
(2000 calibration) portrays a generally more
competitive and more flexible world economy
where, as discussed in the previous section,
(1) trade flows are more evenly distributed
across blocs; (2) trade shares are greater and
home biases in consumption and production are
generally smaller; (3) tradable sectors are larger
in the rest of the world and emerging Asia and
smaller in industrial countries; (4) mark-ups and
nominal rigidities are somewhat lower; and,
(5) the implications of exchange rate shocks for
domestic inflation are smaller on account of
more determined policy efforts to control inflation. Price and income elasticities in trade
equations, however, are identical in the two calibrations, reflecting the unclear net overall effect
of globalization on trade elasticities, as discussed
earlier.
The simulations are illustrative in nature and
should not be interpreted as forecasts by IMF
staff of the resolution of actual imbalances.
Their main purpose is to help explore the
effects of globalization on external adjustment
in the context of imbalances similar to the cur-
39These aspects of financial globalization are modeled broadly along the lines of Hunt and Rebucci (2003) and Ghironi,
·
Işcan, and Rebucci (2005).
40Throughout, size is predicated on the value of nominal GDP in U.S. dollars given the difficulties in constructing world
trade matrices with PPP exchange rates or other weights. Blocs are assumed to grow in real terms at different speeds,
roughly reflecting growth assumptions for these blocs from the latest WEO forecast.
136
THE EFFECTS OF GLOBALIZATION: AN INTEGRATED PERSPECTIVE
rent ones, assuming for simplicity that these are
the result of a few stylized shocks and that policies follow simple, mechanical rules that do not
respond directly to imbalances.41 In this sense,
the experiments below are meant to throw light
on specific issues about the resolution of global
imbalances—such as the implications of alternative financial market conditions for the adjustment—but are by no means an exhaustive
analysis of possible resolution scenarios and
policy options.
How Does Real Globalization Affect
External Adjustment?
A first set of simulations looks at the effects of
real sector globalization on external adjustment
under the assumption that global investors continue to accumulate claims on the United States
for a significant period. In this scenario, which
also involves a moderate fiscal adjustment about
the size proposed by the administration, U.S. net
external liabilities stabilize at about 60 percent
of GDP in the long run, well above current levels, and the U.S. current account deficit falls to a
new long-run level of about 2.5 percent of GDP.
The narrowing of the U.S. current account
deficit is the result of a combination of higher
U.S. real interest rates and higher U.S. saving
ratios—hampering domestic demand—on the
one hand, and a significantly weaker U.S. dollar—boosting net exports—on the other hand.
In the short run, the net effect is a decline in
output growth below trend. During the adjustment, the U.S. trade balance gradually goes into
surplus to meet the additional costs of the
higher long-run net foreign liabilities.
In the other blocs, the rebalancing involves
opposite adjustment patterns. While the U.S.
Figure 3.11. Economies’ Relative Sizes and Trade
Patterns
(Percent of world GDP)
Economies in emerging Asia and the euro area and Japan are relatively larger in
2003 than they were in 1984, whereas the U.S. economy and economies in the rest
of the world have become smaller over time.
1984
United
States
Rest of the
world
1.0
30.9
33.3
0.2
0.5
0.9
1.4
0.5
2.6
2.9
0.5
0.3
0.5
28.7
7.2
0.5
Euro area and
Japan
Emerging
Asia
2003
United
States
Rest of the
world
1.2
30.4
25.6
0.4
0.5
0.9
1.8
0.9
3.2
2.9
0.9
0.4
1.0
34.6
9.4
1.1
Euro area and
Japan
Emerging
Asia
Sources: IMF, International Financial Statistics; IMF, Direction of Trade Statistics; and
IMF staff calculations.
41The simulations assume that in every bloc—excluding
emerging Asia, which in the first set of simulations pegs
its exchange rate to the U.S. dollar—monetary policy is
implemented through an interest rate rule where rates
are set in response to the extent to which inflation deviates from target and output deviates from potential. The
exact specification of this monetary policy rule is
described in Appendix 3.3.
137
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
dollar depreciates, currencies elsewhere appreciate in real terms, and net export growth
declines. At the same time, domestic demand in
the other blocs strengthens, partly reflecting
relatively lower real interest rates and higher
income from larger net foreign assets. In the
long run, the other blocs see their current
account weaken in GDP terms (in the long run,
in emerging Asia the current account to GDP
ratio falls by about 4 percentage points; in Japan
and the euro area it falls by about 1 percentage
point). With the investment income from permanently larger net foreign asset positions, these
blocs can afford to run small long-run trade
deficits.
These fundamental adjustment patterns do
not depend on the state of globalization; they
follow from a country’s long-run external budget
constraint that ties trade balances and net
foreign assets together. However, real sector globalization affects adjustment paths in this rebalancing scenario, as illustrated by the differences
between the rebalancing responses of some key
variables in the low- and high-integration versions
of the model in Figure 3.12. The key results are
as follows.
• With globalization, external adjustment can be
achieved with a smaller real depreciation of
the U.S. dollar and more contained real
appreciations in the other blocs. Globalization
also implies smaller increases in real interest
rates in the United States and in other blocs
during the adjustment.
• As a result, globalization also allows the adjustment to occur with smaller short-run declines
in output growth in all blocs, apart from
emerging Asia.
• The impact of globalization seems largest for
the adjustment in emerging Asia, where output growth during the rebalancing is not as
strong as in the absence of globalization.
The simulation results, therefore, suggest that
real sector globalization broadly facilitates external adjustment. Which aspects of globalization
matter most for this result?
• Direction of trade and trade shares. Trade today is
more evenly distributed across blocs, helping
138
to spread the adjustment in the U.S. current
account deficit more evenly to other world
regions. The positive effects of this, of course,
vary among blocs. Today blocs other than
emerging Asia, for example, are relatively less
affected by negative U.S. developments compared with the 1980s while, at the same time,
they benefit more from positive developments
in emerging Asia, such as the region’s strong
growth. Emerging Asia, however, is now more
internationally exposed with its higher trade
share, so that the slowdown in growth necessary for the U.S. current account to rebalance
has a correspondingly larger impact on this
region compared with the 1980s.
• Economic size of the various blocs. In the 1980s,
the relative economic size of the United States
was bigger, while the economic sizes of emerging Asia, the euro area, and Japan were
smaller, making it harder for them to absorb
the same amount of U.S. assets. Given the
increased economic size of emerging Asia, the
euro area, and Japan in the high-integration
version, an equal-sized increase in U.S. net foreign liabilities (as a percent of U.S. GDP) generates smaller increases in net foreign asset to
GDP ratios in these blocs, other things being
equal. In turn, these entail smaller external
surpluses—that is, relatively smaller savinginvestment balances, which can be attained
through more limited real exchange rate
changes and, especially in the short term, real
interest rate changes, thereby allowing for
higher growth, except in the case of emerging
Asia, mostly for reasons noted above.
In addition, the results depend significantly
on differences in monetary policy strategies across time
(see Box 3.4 for a more detailed discussion of
the links between globalization and monetary
policy). Monetary policies today are more effective and credible, and thus require smaller interest rate changes to stabilize inflation following
demand shocks and during adjustment, partly
because of the lower exchange rate passthrough, as discussed earlier.
In contrast, other key elements, especially
greater flexibility and greater international com-
THE EFFECTS OF GLOBALIZATION: AN INTEGRATED PERSPECTIVE
Figure 3.12. Global Rebalancing Under Benign Financial Market Conditions
(Deviations between baseline responses in GEM 2000 and GEM 1980 calibrations; percentage points; x-axis in calendar quarters, 0 represents 2005Q1)
When financial market conditions are benign, globalization facilitates external adjustment, as this can now be achieved with smaller exchange rate changes and smaller
increases in real interest rates. Under globalization, output losses are also smaller in the short run, except in emerging Asia.
United States
Emerging Asia
Euro Area and Japan
Rest of the World
Output Growth
1
1
1
1
0
0
0
0
-1
-1
-1
-1
-2
-2
-2
-2
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
Trade Balance (percent of GDP)
1
1
1
1
0
0
0
0
-1
-1
-1
-1
-2
-2
-2
-2
-3
-3
-3
-3
-4
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
-4
0
5
10 15 20 25 30 35
-4
0
5
10 15 20 25 30 35
-4
Real Interest Rate
0.5
0.5
0.5
0.5
0.0
0.0
0.0
0.0
-0.5
-0.5
-0.5
-0.5
-1.0
-1.0
-1.0
-1.0
-1.5
-1.5
-1.5
-1.5
-2.0
-2.0
-2.0
-2.0
-2.5
-2.5
-2.5
-2.5
-3.0
-3.0
-3.0
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
-3.0
Real Effective Exchange Rate (+ = depreciation)
8
8
8
8
6
6
6
6
4
4
4
4
2
2
2
2
0
0
0
0
-2
-2
-2
-2
-4
-4
-4
-4
-6
-6
-6
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
-6
Source: IMF staff calculations.
139
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
petition, seem to play a minor role in explaining
the result (see Appendix 3.3 for details).42 This
finding reflects two facts. First, given empirical
evidence discussed earlier, mark-ups and rigidities differ only marginally across the two globalization environments, which is reflected in the
two versions of the model. For example, as
noted in Table 3.8, markups generally fell by less
than 2 percentage points, implying relatively
minor changes in price levels. Second, because
empirical evidence suggests similar relative gains
in flexibility and competition across blocs
(although levels of competition and flexibility
still differ), the related changes in price levels
between the two globalization environments are
similar also and, accordingly, changes in relative
prices between blocs are minor. Hence, while
greater competition and lower rigidities can
facilitate adjustment from a unilateral perspective, they appear to have made less of difference
so far from a global perspective.
What Are the Implications of Changes in
Investor Preferences?
So far, the simulations have assumed—
perhaps rather optimistically—that investors
continue to accumulate U.S. assets for a considerable period. In practice, investor preferences
can be highly volatile, arguably a greater concern in a globalized world where gross and net
external positions are large and where large portions of foreign assets are often purchased and
held by official investors for motives other than
pure portfolio optimization. To explore the
potential impact of a change in investor preferences, two alternative scenarios are considered,
using the 2000 calibration of GEM.
• In the first scenario, investors in the euro
area, Japan, and the rest of the world are
unwilling to continue accumulating U.S. assets
and begin to gradually reduce their desired
holdings of U.S. assets (back to 2001 levels by
2010). However, investment behavior in
emerging Asia does not change, and currencies in the region remain closely linked to the
U.S. dollar.
• In the second scenario, it is assumed that all
investors outside the United States, including
emerging Asia, gradually reduce their desired
holdings of U.S. assets (back to 2001 levels by
2010). This scenario also assumes—for illustrative purposes only—that emerging Asia moves
immediately to a floating exchange rate
regime, with monetary policy determined by
an interest rate rule broadly similar to that
assumed for the other blocs.
Such changes in investor preferences clearly
complicate global rebalancing. Compared with
the previous scenario (2000 calibration), the key
results are as follows (see Figure 3.13).
• Capital flows to the United States slow sharply,
requiring a more abrupt adjustment in the
U.S. current account. Correspondingly, U.S.
interest rates rise relative to the baseline scenario, the U.S. dollar has to depreciate sooner
and more sharply, and U.S. output slows more
markedly. Unsurprisingly, all these effects are
larger in the case where demand for U.S.
assets falls in all regions.
• The depreciation of the U.S. dollar is matched
by an appreciation of exchange rates—and
lower trade and current account balances—in
blocs with flexible exchange rates (Japan, the
euro area, and the rest of the world). In
emerging Asia, however, developments vary
substantially across the two scenarios. In the
former, the real exchange rate appreciates relatively less as the Asian currencies move down
with the U.S. dollar. In contrast, when Asian
currencies float, the exchange rate appreciates
sharply. In this latter case, the appreciation in
the other two blocs is somewhat less than
under the first scenario, as the decline in
42Greater flexibility, competition, and lower pass-through (through a more credible monetary policy) generally tend to
facilitate external adjustment in the sense that they push key variables in the direction required by the adjustment. The
exception is emerging Asia, where, contrary to other blocs, monetary policy is assumed to have stayed the same between
the 1980s and the 2000s.
140
THE EFFECTS OF GLOBALIZATION: AN INTEGRATED PERSPECTIVE
Figure 3.13. Global Rebalancing Under Adverse Financial Market Conditions
(Shock minus control, where "shock" is adverse scenario and "control" is 2000 baseline under benign scenario; percentage points; x-axis in calendar
quarter, 0 represents 2005Q1)
The effects of globalization when financial market conditions are adverse are more mixed. As capital flows to the United States dry up, the United States can only rebalance its
external position through a sharper contraction and a bigger U.S. dollar depreciation. Other blocs largely benefit, as their saving need not rise to pay for additional purchases of
U.S. assets.
Investor preference shocks everywhere/emerging Asia floats its currency
Investor preference shocks in Japan/euro area and rest of the world only
United States
Emerging Asia
Euro Area and Japan
Rest of the World
Output Growth
2
2
2
2
1
1
1
1
0
0
0
0
-1
-1
-1
-1
-2
0
5
10 15 20 25 30 35
0
5
-2
10 15 20 25 30 35
0
5
10 15 20 25 30 35
-2
0
5
10 15 20 25 30 35
-2
Trade Balance (percent of GDP)
3
3
3
3
2
2
2
2
1
1
1
1
0
0
0
0
-1
-1
-1
-1
-2
-2
-2
-2
-3
-3
-3
0
5
10 15 20 25 30 35
0
5 10 15 20 25 30 35
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
-3
Real Interest Rate
1.0
1
1
1
0.0
0
0
0
-1.0
-1
-1
-1
-2.0
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
-2
0
5
10 15 20 25 30 35
-2
0
5
10 15 20 25 30 35
-2
Real Effective Exchange Rate (+ = depreciation)
16
16
16
16
12
12
12
12
8
8
8
8
4
4
4
4
0
0
0
0
-4
-4
-4
-4
-8
0
5
10 15 20 25 30 35
0
5
10 15 20 25 30 35
-8
0
5
10 15 20 25 30 35
-8
0
5
10 15 20 25 30 35
-8
Source: IMF staff calculations.
141
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
emerging Asia’s desire for foreign assets is
accompanied by relatively faster domestic
demand growth, which, unlike in the first scenario, contributes to the absorption of higher
U.S. net exports.
• In both scenarios, real interest rates outside
the United States fall, and—despite slower
U.S. growth—GDP growth in the rest of the
world rises moderately (or in emerging Asia
stays broadly flat relative to the baseline in the
first scenario). This rather benign outcome
partly reflects the specification of the shift in
investor preferences, as the decline in desired
asset holdings in the rest of the world is
accompanied by a reduction in desired savings, which boosts consumption.43 In practice,
demand outside the United States could fail to
pick up—for example, because of adverse confidence effects—and GDP growth would correspondingly be weaker.
Overall, the simulations underscore the potential risks attendant on a reversal of investor preferences and of an abrupt, as opposed to a
gradual, change in central bank behavior in
Asia. Moreover, while the simulations do not
take full account of some mitigating factors—
notably positive valuation effects for the United
States from a depreciating U.S. dollar, as discussed earlier44—on balance they probably
underestimate the risks for a number of reasons.
First, as noted above, growth in the rest of the
world could well be weaker than suggested by
the model. Second, large exchange rate changes
may induce financial market turbulence, especially since investors may not always have the
perfect foresight assumed in the model. Finally,
GEM does not fully incorporate the effects of
higher risk premia (since it only distinguishes
net foreign asset positions, not the underlying
gross asset and liability positions). At current levels of U.S. external liabilities, a permanent 100
basis points increase in the risk premium on
U.S. assets raises the required long-run trade surplus in the United States by 0.7 percentage point
of GDP (Table 3.9), which would require a substantially larger depreciation of the U.S. dollar
(and, of course, greater appreciations elsewhere). In contrast, the risk premium effect is
less than half this size in the less globalized
world of the 1980s, highlighting the importance
of the buildup of gross assets and liabilities in
recent years.
Summary and Policy Implications
Against the background of concerns about
current global imbalances, the multidimensional
relationship among globalization, external
imbalances, and international adjustment has
become of increasing interest to policymakers.
Recent empirical evidence and theoretical considerations alike suggest that financial globalization has contributed to an environment in which
large current account surpluses or deficits
emerge and are sustained, partly reflecting the
revealed willingness of investors to hold larger
shares of foreign assets in their portfolios. Real
sector globalization has clearly reshaped the
magnitudes, composition, and direction of trade
flows in the global economy. However, empirical
evidence to date is ambiguous as to whether this
has increased the sensitivity of the demand for
traded goods and services to changes in relative
prices or demand conditions.
As the simulations and related analysis above
show, real and financial globalization should
43Simulation results from scenarios where the shift in investor preferences is reflected in a persistently higher risk premium on U.S. assets rather than in changes in desired net foreign asset positions suggest similar growth patterns during
adjustment. In particular, the increase in the risk premium on U.S. assets leads to relatively lower real interest rates elsewhere, as capital leaves the United States and is invested elsewhere. This provides a boost to domestic demand that more
than offsets the adverse growth impact of lower net export growth.
44In GEM, there is only one international bond, which is denominated in U.S. dollars. Hence, while emerging Asia, the
euro area, Japan, and the rest of the world experience adverse valuation shocks when the U.S. dollar depreciates, there are
no gains for the United States. Moreover, in GEM the valuation effects arise only on net foreign asset positions and not on
gross positions.
142
SUMMARY AND POLICY IMPLICATIONS
Table 3.9. Impact of Risk Premium Shocks on Long-Run Trade Balances1
Changes in Long-Run Trade Balances (percent of GDP)
Gross Foreign Assets
(percent of GDP)
After a Risk Premium Shock (in bps)2
______________________________________________________
ρ = 100
ρ = 200
ρ = 300
ρ = 400
United States
Low integration (1984)
High integration (2003)
26.6
70.5
0.3
0.7
0.5
1.4
0.8
2.1
1.0
2.8
Emerging Asia
Low integration (1984)
High integration (2003)
12.4
56.7
0.1
0.6
0.2
1.1
0.4
1.7
0.5
2.2
Japan and Europe
Low integration (1984)
High integration (2003)
35.9
147.8
0.4
1.4
0.7
2.9
1.1
4.3
1.4
5.8
1In the long run, sustainable trade balances (i.e., trade balances consistent with stable ratio of net foreign assets to GDP) are obtained by
adding to the long-run trade balances from GEM a term that captures the effect on the trade balance of changes in the value of gross foreign
assets. This, in turn, depends proportionately on the interest rate differential (r N) and the risk premium on those assets (ρ), as shown in (1)
below:
rN – ρ
tb = tbGEM – ––––– a–,
1+g
(
)
(1)
where a is gross foreign assets (as a share of GDP) and g is the rate of output growth (see Appendix 3.1 for details). The gross asset positions
used to calculate the trade balances in the two scenarios are the historical values reported in the table. In both scenarios, net foreign returns r N
are calculated as the averages for the period 1990–2003, and the long-run rate of output growth is taken to be 2 percent.
2Results reported in the table show how long-run trade balances change with different values for the risk premium in the two scenarios. For
example, if net foreign returns for the United States decline (rise) by 400 bps in the high integration scenario, the trade balance required to stabilize U.S. net foreign assets increases (falls) by 2.8 percentage points of GDP. By contrast, in the low integration scenario, a similar fall in net foreign returns increases the long-run trade balance by 1 percent of GDP.
generally facilitate global rebalancing, provided
financial conditions remain benign. With output
levels in major economic blocs now more equal
in size, the financing of the U.S. deficit requires
relatively smaller surpluses elsewhere; the more
even distribution of trade flows across the world
contributes to better burden sharing among
countries; and more credible monetary policies
reduce the output cost of adjusting to shocks.
That said, globalization has fundamentally
changed neither the nature of adjustment nor
the magnitudes involved: ultimately exchange
rates and trade balances will need to adjust, and
adjust substantially.
The analysis also suggests that while home
bias in investor preferences and restrictions to
net international borrowing have decreased,
they have not disappeared. Accordingly, the
accommodation of further increases in U.S. net
external liabilities should not be taken for
granted. If financial market conditions prove to
be less benign, with only limited or perhaps
temporary scope for further increases in U.S.
net external liabilities, global rebalancing is
likely to involve greater risks. In particular,
initial changes in real exchange and interest
rates would be much larger, and the short-term
output costs of a global rebalancing clearly
higher, particularly in the United States. In an
environment where gross assets and liabilities
are substantially higher than they were in the
past—a feature not fully captured in the
simulations—the potential for large changes
in exchange rates, interest rates, and risk premia to lead to disruptive financial market
turbulence in the short term must be correspondingly elevated.
In sum, the central message of the chapter is
that policy makers need to take advantage of the
scope that globalization provides to facilitate
adjustment, while remaining mindful that this
adjustment must eventually take place, and of
the potential risks associated with unexpected
changes in investor preferences along the way.
These simultaneous objectives would appear
most likely to be achieved by the adoption of a
credible medium-term policy framework consistent with achieving external and internal
143
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
balance within a reasonable period of time,
thereby anchoring expectations and reducing
risks of reversals in investor preferences. This
underscores the importance of implementing
the cooperative strategy to address imbalances
recommended by IMF Governors at the last
IMFC meeting in October 2004. Unfortunately,
as discussed in Chapter I, to date only moderate
progress has been made.
Appendix 3.1. Net Foreign Assets,
Valuation, and Adjustment:
A Glossary of Terms
The main author of this appendix is Roberto
Cardarelli.
Accounting for Changes in Net Foreign Assets
Changes in a country’s net foreign asset position (NFA) are, by definition, described by the
following accounting identity:
NFAt – NFAt–1 ≡ NXt + CTt + IAt + KAt + EOt + KGt ,
where NX denotes net exports of goods and services; CT, current transfers; IA, the investment
income balance (with the sum of the three
being the current account balance); KA, capital
transfers; EO, errors and omissions; and KG, net
capital gains (losses if negative).
Valuation Effects
Using lowercase letters to show net foreign
assets and net exports as a share of GDP, excluding current and capital transfers on account of
their relatively small size, ignoring errors and
omissions, and assuming that returns on foreign
assets are equal to returns on foreign liabilities,
the accounting identity can be expressed as
follows:
it – γt
kt
nfat – nfat–1≡ nxt + –––––
nfat–1 + –––––
nfat–1,
1 + γt
1 + γt
where γ is the country’s rate of nominal growth,
i is the yield rate, and k is the rate of capital
gain on external assets and liabilities. The last
term on the right-hand side defines the valua-
144
tion effect shown in figures and tables in this
chapter.
External Solvency
Because the accounting identity for net foreign assets holds in every period, it can be
expressed on a cumulative basis—the so-called
intertemporal budget constraint for a country—
leading to the well-known proposition that a
debtor country is solvent if the net present value
of its future net exports equals the current
period value of its net foreign liabilities.
External Sustainability and Imbalances
The intertemporal solvency condition serves
only to exclude cases in which a country
indulges in perpetual debt refinancing of its
external imbalances (a so-called Ponzi game),
but is otherwise consistent with many patterns
for future net foreign assets, including one in
which they increase exponentially at a rate
lower than the difference between the yield on
net external liabilities and the growth rate. A
more operational definition of sustainable net
exports consists in imposing that they stabilize
the ratio of net foreign assets to GDP to a certain level. This could be its current level, or a
level that is considered consistent with the international investors’ willingness to lend to the
country (which, in turn, may relate to their perception of the country’s willingness and ability
to meet its external obligations), or a level consistent with some optimal, theoretically predicted, benchmark. In this case, the level of net
exports that a country needs to sustain in the
long run to keep the ratio of net foreign assets
–––
to GDP constant at a level denoted with nfa can
be written as
r – g –––
nx t* = ––––– nfa ,
1+g
where r is the total real return (yield plus capital
gains) on net foreign assets and g is the country’s rate of real growth. In this framework, the
difference between the current level of net
APPENDIX 3.2. ECONOMETRIC EVIDENCE ON THE VALUATION CHANNEL
exports and nx* can be interpreted as a measure
of the degree of external imbalances.
the role of the valuation channel in external
adjustment.
Financial Integration and External Adjustment
Modeling and Econometric Strategy
Introducing different rates of return on foreign assets and liabilities, the accumulation
identity for net foreign assets can be written as
The role of the valuation channel in the
process of adjustment of external imbalances is
assessed empirically by examining the dynamic
responses of net exports and net foreign assets
to shocks that move the two variables away from
their long-run relationship implied by a country’s intertemporal budget constraint. The latter
is given by
r tL – gt
r tA – r tL
nfat – nfat–1≡ nxt + ––––––
nfat–1 + ––––––
a ,
1 + gt t–1
1 + gt
where r L is the total real return (yield plus capital gains) on foreign liabilities, r A is the total real
return on foreign assets, and a denotes gross foreign assets (as a share of GDP). Sustainable net
export levels can then be defined as
r L – g ––– r A – r L
nx* = –––––– nfa – –––––– a–.
1+g
1+g
This expression shows that a larger (steady-state)
stock of gross foreign assets imposes a bigger
trade adjustment if the return differential r A – r L
is negative, highlighting the greater exposure
arising from larger external positions. A numerical example of this effect is shown in Table 3.9,
where the steady-state net exports derived from
GEM are equal to the first term on the righthand side of the expression above. Sustainable
net export levels after a risk premium shock—
modeled as increases in r L while keeping r A
unchanged—are then obtained by adding the
second term, taking the level in the year underlying the calibration as the steady-state ratio of
gross foreign assets to GDP a–.
Appendix 3.2. Econometric Evidence on
the Valuation Channel
The main author of this appendix is Roberto
Cardarelli.
This appendix provides details on the econometric evidence discussed in the main text about
∞
NFAt = ∑ Rt,t+i(Xt+i – Mt+i ),
t=0
(1)
where NFAt represents the real (based on GDP
deflator) level of net foreign assets; Xt and Mt ,
the volumes of export and imports of goods and
services, and Rt,s—the discount factor for period
s net exports—is a function of the real returns
on net foreign assets rt (see also Appendix 3.1).45
Following Gourinchas and Rey (2004) and
Corsetti and Konstantinou (2004), it is possible
to derive the following log-linear approximation
of equation (1):
xt – γmt + (γ – 1)nfat
∞
= ∑ ρi[∆xt+i – γ∆mt+i + (γ – 1)rt+i ],
i=1
(2)
where lowercase letters denote logarithms.46
Under the assumptions that rt, ∆xt, and ∆mt are
stationary, expression (2) implies that x, m, and
nfa should be cointegrated, and the left-hand
side (the cointegration residual) represents deviations from the long-run relationship among the
three variables. Expression (2) also implies that
if the cointegrating residual is not a constant, it
has to predict either changes in the future net
exports or in net foreign returns or both. For
example, if a country develops a larger trade
deficit relative to what is consistent with its
steady-state net foreign assets position, the deviation from the long-run relationship has to be
s
1
particular, Rt,s = ∏ ––––––
.
i=0 (1 + rt+i)
46In expression (2), γ and ρ are constants related to the log-linearization of the country’s intertemporal budget constraint (see Corsetti and Kostantinou, 2004).
45In
145
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
corrected either by an increase in future net
exports or by a future increase in net returns on
the stock of its foreign assets and liabilities, or a
combination of the two.
Based on this framework, the relevance of a
valuation channel in the process of adjustment
of external imbalances is assessed in two different ways.
• Using annual data from 1970 to 2003,47 the
following vector error correction model
(VECM) representation of the three variables
Y = [x, m, nfa] was estimated for 21 industrial
countries and 28 emerging market countries:48
∆Yt = c + αβ′Yt–1 + Γ(L)∆Yt–1 + et ,
where ∆Yt is the first difference of the vector
Yt; Γ is a coefficient matrix, β′Yt–1 is the last
period’s deviation from the long-run relationship (cointegrating residual in period t – 1), L
is the lag operator, and α is a vector of adjustment coefficients that determines how each
variable adjusts to restore the long-run relationship if the cointegration residual signals a
deviation from that relationship.
If the coefficient α in the equation for net
foreign assets is statistically significant, this
implies that changes in net foreign assets play
a role in the adjustment to restore the longrun relationship among the three variables x,
m, and nfa (see Lettau and Ludvigson, 2004,
and Corsetti and Konstantinou, 2004). In fact,
that role varies with the value of α. The larger
the value of α in the equation for net foreign
assets, the bigger the role played by changes in
that variable in the overall adjustment. Moreover, using cointegration results (and assuming orthogonality between shocks), one can
distinguish between permanent and transitory
shocks on the variables in Y.49 In particular, it
has been shown that variables that participate
significantly in the adjustment are assigned a
relatively large (small) weight in the transitory
(permanent) innovations. A large value of α
in the equation for net foreign assets, therefore, implies that a significant share of the
variation in net foreign assets is explained by
transitory movements. These, in turn, can be
interpreted as exchange-rate- or asset-pricerelated valuation effects that force the stock of
net foreign assets to temporarily deviate from
its trend to help restore the long-run relationship with net exports.
If, on the other hand, the coefficient α in
the equation for net foreign assets is not significantly different from zero, net foreign assets
do not play a role in the adjustment process. In
this case, the adjustment has to come exclusively from trade flows and any shock has permanent effects on net foreign assets, which will
follow a random walk process. This, in turn, is
interpreted as evidence against the notion that
the valuation channel facilitates the adjustment
of external imbalances.
• Using quarterly data on net foreign assets and
returns, the implication that deviations from
the long-run relationship may predict future
changes in net foreign returns was formally
tested for G-7 countries. In particular, insample forecasts were performed by regressing
the changes between t and t + k of export,
import, net foreign returns, and real exchange
rates on the estimated cointegrating residuals
(see Gourinchas and Rey, 2004). For example,
the regression for export growth between t
and t + k ahead takes the form
Xt+k – Xt = c + δ(β̂′Yt–1) + et .
47For some countries, the sample period is slightly shorter, especially for emerging market countries, where data tend to
start from the mid-1970s.
48The industrial countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Iceland, Ireland,
Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the
United States. The emerging market countries are Argentina, Brazil, Chile, People’s Republic of China, Colombia, Costa
Rica, Côte d’Ivoire, Ecuador, Egypt, Hungary, India, Indonesia, Jordan, Korea, Malaysia, Mexico, Morocco, Pakistan, Peru,
the Philippines, Poland, Singapore, South Africa, Taiwan Province of China, Thailand, Turkey, Uruguay, and Venezuela.
49Methodologies to decompose transitory and permanent components based on the cointegration results have been
described by King and others (1991), Gonzalo and Granger (1995), Gonzalo and Ng (2001), and Warne (1993).
146
APPENDIX 3.2. ECONOMETRIC EVIDENCE ON THE VALUATION CHANNEL
Results
For each of the 49 countries considered, the
existence of cointegration among the logarithm
of real exports, imports, and net foreign assets
was tested using Johansen’s trace test.50 For the
vast majority of countries (37), the results supported the null hypothesis of one cointegrating
relationship among the three variables at a
5 percent confidence level. However, since the
intertemporal budget constraint is an identity
that needs to hold in the long run, cointegration was imposed for all countries in the
subsequent analysis, as was the condition that
the cointegrating vector is of the form β =
[1, –γ, (γ – 1)] (formal tests suggested rejection
of that restriction in about one-half of the
countries).
The subsequent estimates of the VECM show
that the adjustment coefficient α in the equation
for ∆nfat was significantly different from zero for
a number of industrial countries (8 out of 21),
including Japan and the United States, for which
about one-half of the variation in net foreign
assets was explained by transitory shocks. On the
contrary, the adjustment coefficient in the equation of net foreign assets was statistically significant for only 4 of the 28 emerging market and
developing countries (net foreign assets were
found weakly exogenous to the system for all
other countries).
The difference between the two sets of countries emerges from Table 3.10, showing the
arithmetic average of the countries’ variance
decompositions (mean group estimator).51 On
average, the share of the variation in net foreign
assets explained by transitory shocks at a oneyear horizon is about twice for the 21 industrial
countries as much as that for the 28 emerging
market countries considered.
The ability of deviations from trend to predict
future changes in net foreign returns was
Table 3.10. Panel VECM: Forecast Variance
Decomposition1
(Percent)
Industrial Countries
__________________________________________
X
M
NFA
__________
__________
__________
Horizon
P
T
P
T
P
T
1
2
3
4
10
86
87
89
89
95
14
13
11
11
5
70
74
79
84
92
30
26
21
16
8
79
81
83
85
90
21
19
17
15
10
Emerging Markets
__________________________________________
X
M
NFA
__________
__________
__________
Horizon
P
T
P
T
P
T
1
2
3
4
10
82
85
87
90
96
18
15
13
10
4
48
58
65
70
85
52
42
35
30
15
90
93
95
96
99
10
7
5
4
1
Source: IMF staff calculations.
1Fraction of the variance of the h-year ahead forecasts error of
exports (X), imports (M), and net foreign assets (NFA) that is
explained by the two permanent (P) shocks combined and the
single transitory (T ) shock. Values reported are the averages across
countries.
assessed through in-sample forecasts only for
the G-7 countries, for which quarterly series on
net foreign assets and net foreign returns were
constructed (see below). Evidence of cointegration was found only for Canada, Germany,
Japan, and the United States, so Table 3.11
shows the results only for these countries. In
all four countries, cointegrating residuals have
substantial predictive power (reflected in a
high relative value of the R 2 of the regressions)
for export and/or import growth over relatively
long horizons. The coefficients generally have
the expected sign (negative for export growth
and positive for import growth), and tend to
be statistically significant for long horizons. By
contrast, cointegrating residuals do not seem
to have the same general ability to predict
future net foreign returns. In these regressions,
50The number of lags in the VECM specification was chosen in accordance with standard lag order selection criteria
(such as the Akaike and Hannan-Quinn information criteria). However, given the limited number of time-series observations available, a more parsimonious choice was often preferred, as long as it resulted in normally distributed and serially
uncorrelated errors.
51See Pesaran and Smith (1995) and Rebucci (2003), for a description of the properties of this estimator.
147
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Table 3.11. Predictive Ability of Cointegrating Residuals
Forecast Horizon (quarter)
_________________________________________________________________________________________________________
United States
Japan
___________________________________________________
__________________________________________________
1
2
4
8
12
24
1
2
Real Export Growth
R2
0.00
0.00
0.00
0.00
–0.02
0.01
–0.05
0.02
–0.12
0.03
–0.38
0.13
–0.07
0.08
–0.16
0.16
Real Import Growth
R2
0.00
0.00
0.00
0.00
R2
–0.02
0.04
–0.02
0.07
0.01
0.00
0.11
0.04
–0.01
0.06
0.24
0.12
0.24
0.05
–0.01
0.00
–0.02
0.00
–0.01
0.02
0.00
0.00
0.01
0.00
0.01
0.00
0.02
0.04
0.04
0.05
0.08
0.09
0.15
0.11
2
–0.06
0.02
–0.12
0.04
4
8
0.20
0.08
0.26
0.04
0.06
0.01
0.11
0.02
R2
–0.30
0.08
24
1
2
–0.39
0.11
–0.90
0.41
0.00
0.00
0.02
0.00
0.01
0.00
0.01
0.00
0.08
0.02
0.21
0.05
0.03
0.01
0.03
0.01
0.03
0.01
0.03
0.03
0.17
0.02
0.18
0.02
0.10
0.07
0.00
0.00
0.01
0.00
0.03
0.00
0.12
0.02
–0.59
0.53
0.00
0.00
0.26
0.06
0.63
0.23
1.14
0.56
0.00
0.00
–0.01
0.01
–0.02
0.02
–0.01
0.01
0.25
0.04
0.64
0.15
0.81
0.16
0.71
0.12
4
8
12
24
–0.01
0.00
–0.23
0.04
–0.51
0.16
–0.59
0.23
0.22
0.13
0.42
0.19
0.46
0.14
0.45
0.12
0.26
0.03
0.07
0.12
0.04
0.10
Real Net Foreign Return
0.02
0.01
–0.02
0.02
–0.02
0.00
Real Effective Exchange Rate (+ = appreciation)
R2
–0.43
0.32
Real Import Growth
Real Net Foreign Return
R2
–0.50
0.39
Real Export Growth
Real Import Growth
R2
–0.35
0.32
Canada
__________________________________________________
12
Real Export Growth
–0.19
0.05
24
Real Effective Exchange Rate (+ = appreciation)
Germany
___________________________________________________
1
12
Real Net Foreign Return
Real Effective Exchange Rate (+ = appreciation)
R2
8
Real Import Growth
Real Net Foreign Return
–0.02
0.08
4
Real Export Growth
0.18
0.06
–0.02
0.00
0.00
0.00
0.05
0.04
Real Effective Exchange Rate (+ = appreciation)
0.51
0.24
0.02
0.00
0.04
0.01
0.07
0.01
0.14
0.02
0.36
0.10
0.46
0.15
Source: IMF staff calculations.
Note: For each variable, the table shows (1) the coefficient δ of regressions of the form: Yt+k – Yt = c + δ(β̂Yt–1) + et, where the left-hand side
denotes the change in the variable between t + k and t, and the independent variable is the deviation of net exports and net foreign assets from
their long-run relationship (cointegration residual) and (2) the R 2 of the regression. Bold values denote coefficients that are significant at a 5 percent level (based on Newey-West robust standard errors).
the coefficients have the expected negative
sign and are statistically significant only for
the United States. Finally, deviations from
trend seem to have some predictive power for
real exchange rates only for medium and
long horizons, when coefficients are generally
found to be statistically significant (the United
States again being the exception). This is consistent with the adjustment taking place more
through long-term (exchange-rate-related)
realignments of trade flows, rather than
148
through short-term valuation changes in the
stock of foreign assets.
Data
Exports and imports: Annual and quarterly data
from the OECD (Economic Outlook).
Net foreign assets: Annual data from Lane
and Milesi-Ferretti (2005b). Quarterly data
on the stocks of foreign assets and liabilities
were obtained by interpolating the Lane and
APPENDIX 3.3. DETAILS ON RIGIDITIES AND MONETARY POLICY RULE USED IN THE GEM SIMULATION
Milesi-Ferretti annual data, using the quarterly
path of capital inflows and outflows (from the
IMF’s International Financial Statistics).
Quarterly returns on foreign assets and liabilities:
Following the methodology adopted by
Gourinchas and Rey (2004) to estimate the
returns on foreign assets and liabilities for the
United States, quarterly rates of return of foreign assets and liabilities were constructed for
the G-7 countries as weighted averages of the
returns on four different subcategories of the
financial accounts (equity, FDI, debt, and other
assets and liabilities), with weights given by their
relative shares in the total stock of assets and
liabilities.
Returns were estimated as follows. For equity
assets, returns were calculated as the weighted
average of quarterly (dollar) returns on the
Morgan Stanley Capital International (MSCI)
indexes for the major countries of nonresident
issuers, with weights from the geographical allocation of foreign equity assets as reported in the
IMF’s 2001 Coordinated Portfolio Investment Survey
(CPIS). For FDI assets, the same returns were
used but the weights reflected the geographical
allocation of FDI assets, as reported by the
OECD’s International Direct Investment database.
For debt assets, the returns were calculated as a
weighted average of the (dollar) long-term
interest rates on government bonds (from the
OECD’s Economic Outlook), with weights derived
from the geographical allocation of long-term
debt assets reported in the 2001 CPIS. For other
assets, the returns were calculated as a weighted
average of the (dollar) short-term interest rates
(from the OECD’s, Economic Outlook), with
weights from the 2001 CPIS foreign geographical allocation of short-term debt assets. For each
country, the returns on equity and FDI liabilities
were estimated as the return on their MSCI
indexes (in U.S. dollars), while returns on debt
and other liabilities were estimated as the
national long- and short-term interest rates (in
U.S. dollars). Conversion into U.S. dollars was
obtained using end-of-period exchange rates
from the OECD’s Main Economic Indicators.
Nominal returns were then deflated using the
U.S. GDP deflators, and converted into local
currencies using bilateral real exchange rates
against the U.S. dollar.
Appendix 3.3. Details on Rigidities and
Monetary Policy Rule Used in the
GEM Simulation
The author of this appendix is Nicoletta Batini.
Preliminary Estimates of Nominal and
Real Rigidities
To examine whether nominal and real rigidities have changed in the past three decades, IMF
staff estimated structural equations for inflation
using data from G-7 countries. These equations,
known in the literature as New Keynesian
Phillips Curves (NKPC) (see Sbordone, 2005),
describe the evolution of aggregate inflation
with respect to marginal costs faced by firms in
the production of goods and services, and capture the fact that prices are set in a forwardlooking way. The extent of nominal and real
rigidities can be gauged by looking at reducedform parameter estimates that, in turn, depend
nonlinearly on structural parameters in the
equations.
For the standard specification of the NKPC, it
is assumed that steady-state inflation is zero, and
that the world is “hybrid” in the sense that a fraction of firms are forward-looking and set prices
in a staggered fashion as in Calvo (1983) while
the rest are myopic and set their prices equal to
the average price in the previous period. The
NKPC can then be written as follows (variables
are expressed in log deviation from their steadystate values):
πt = λmct + γf Et {πt+1} + γbπt–1,
(1)
where πt is inflation, mct is the real marginal cost,
and Et is the expectational operator based on
information at time t. The marginal cost is unobservable and, thus, it is usually proxied by an
appropriate measure of the unit labor cost. The
parameters λ, γf , and γb are functions of the
probability faced by firms of resetting their
149
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Table 3.12. Estimates of Nominal and Real
Rigidities in an Earlier and a Later Sample
Degree of
Nominal
Rigidity (D)
_________________
Earlier
Later
sample
sample
Canada
France
Germany
Italy
Japan
United Kingdom
United States
9.92
2.31
3.12
1.47
10.45
2.81
3.41
3.12
12.50
3.84
11.00
9.00
2.17
2.86
Degree of
Real Rigidity (λ)
________________
Earlier
sample
Later
sample
0.25
0.01
0.25
0.45
0.009
0.23
0.16
0.09
0.00
0.18
0.06
0.009
0.00
0.00
Source: IMF staff calculations.
prices in each period (1 – θ), the fraction of
myopic firms (ω), and the discount rate (β).
The degree of nominal rigidity, or price inertia, is typically captured by D = (1/1 – θ)(1/1 – ω),
which measures the average duration of price
contracts when a fraction of firms are myopic
(see Galí, Gertler, and López-Salido, 2001). The
degree of real rigidity is usually captured by λ,
where λ ≡ (1 – ω)(1 – θ)(1 – βθ)φ–1, and where,
in turn, φ ≡ θ + ω[1 – θ(1 – β)] (see Coenen and
Levin, 2004).
Table 3.12 reports values of D and λ from preliminary results obtained by estimating equation
(1) for a group of G-7 countries with the
General Method of Moments, using quarterly
data over two different samples: an earlier sample covering the 1970s up to the mid-1980s, and
a later sample covering the second half of the
1980s to the present.52 Results indicate that real
rigidities seem to have diminished over time,
while the direction of nominal rigidities has
been more mixed, especially in the case of
European countries, in line with findings in
Batini (2002) and Khan (2004).
Monetary Policy Rule Used in the Simulations
The simulations shown in the chapter use the
same specification for the monetary policy rule
for the interest rate as that of a policy rule esti-
52Exact
150
mated by Orphanides (2003) for the United
States. The properties of inflation-forecast-based
rules like this, with and without output terms,
are discussed in Batini and Haldane (1999). The
exact specification used in the chapter is as follows:
a + θ ∆α y
it = θ0 + θiit–1 + θπit+3
∆y
t+3 + θy yt–1,
(2)
where it is the short-term nominal interest rate;
a = [(p
T
πt+3
t+3 – pt+1) – π t is the one-year-ahead
forecast of inflation starting at t – 1; yt–1 = qt–1 –
q*t–1 is the output gap in period t – 1; and ∆α yt+3 =
yt+3 – yt+1 = ∆α qt+3 – ∆α q*t+3 is the one-year-ahead
growth forecast relative to potential growth. In
the model simulations, price inflation used in
the rule is taken to be consumer price inflation.
πTt is the annual inflation target. This is set to
2!/2 percent for the United States and 2 percent
for Japan and the euro area and the rest of the
world. In the simulation with adverse global
financial market conditions (Figure 3.13),
emerging Asia is assumed to float its currency
and move to a rule like (2), with an inflation target that is gradually reduced to 2!/2 percent by
2007. Details of the remaining parameters used
in the rule in the two vintage model calibrations
of GEM are given in Table 3.8 in the main text.
Decomposition of Differences in Two
GEM Calibrations
Table 3.13 shows a breakdown of the contribution by various groups of parameters to the differences in the response paths of output growth,
inflation and interest rates, and real exchange
rates at various horizons between the less globalized and more globalized environment. The
table illustrates that while the economic size of
various blocs, a more equal distribution of trade,
and differences in monetary policy are key in
explaining differences across simulations
between the “low integration” and the “high
integration” worlds, greater flexibility and
greater international competition seem to play a
sample periods are slightly different for each country to maximize individual fits.
APPENDIX 3.3. DETAILS ON RIGIDITIES AND MONETARY POLICY RULE USED IN THE GEM SIMULATION
Table 3.13. Decomposition of Differences in Response Paths by Parameter Groups
(Differences in percentage points)
Total
Trade
Markups
Flexibility
Contracting1
Monetary
Policies
Cross Effects
Decomposition for emerging Asia
First two years
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
–2.11
0.16
–0.44
0.82
–2.46
–0.74
0.92
0.82
0.02
–0.03
0.24
0.06
–0.26
–0.30
1.28
0.39
0.28
1.15
–2.49
–0.13
0.30
0.08
–0.39
–0.32
First five years
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
–1.46
–0.74
1.85
2.72
–1.56
–1.34
2.97
3.27
0.01
–0.05
0.19
0.19
–0.00
–0.06
0.31
0.31
0.01
0.70
–1.58
–0.36
0.09
0.02
–0.05
–0.68
Long run
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
–0.02
–0.01
–0.05
6.04
–0.03
–0.01
–0.03
7.79
–0.00
0.00
–0.00
0.69
0.02
0.01
–0.01
–0.21
0.00
0.00
–0.01
–1.12
–0.01
–0.00
0.01
–1.12
Decomposition for Japan/euro area
First two years
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
0.07
–0.09
0.20
1.44
0.06
–0.08
0.18
1.56
–0.01
–0.01
0.02
–0.03
–0.00
0.01
–0.01
–0.13
0.05
0.01
–0.06
–0.09
–0.03
–0.02
0.06
0.13
First five years
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
0.03
–0.09
0.20
2.31
0.05
–0.08
0.21
2.29
–0.01
–0.00
0.01
–0.06
–0.02
0.01
–0.04
–0.07
0.00
–0.01
0.01
–0.05
0.00
–0.00
0.02
0.20
Long run
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
–0.00
–0.00
–0.02
3.55
–0.00
–0.00
–0.01
3.04
0.00
–0.00
0.00
–0.12
0.00
–0.00
–0.00
0.08
0.00
0.00
–0.01
0.30
–0.00
–0.00
0.00
0.25
Decomposition for the United States
First two years
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
0.66
0.97
–2.93
–0.57
0.11
0.11
–0.33
–0.55
0.00
–0.01
0.00
–0.03
0.02
0.01
0.05
–0.12
0.50
1.01
–3.04
0.02
0.02
–0.15
0.39
0.12
First five years
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
0.18
0.54
–1.69
–2.25
0.09
0.04
–0.14
–1.90
0.01
–0.00
0.01
–0.05
0.03
–0.04
0.13
–0.14
0.08
0.68
–2.04
–0.26
–0.02
–0.13
0.35
0.10
Long run
Output growth
Year-on-year inflation
Real short-term interest rate
Real effective exchange rate
0.01
0.00
–0.12
–4.59
0.01
0.00
–0.11
–2.92
–0.00
0.00
–0.00
–0.20
0.00
0.00
–0.00
–0.17
0.00
–0.00
–0.00
–1.16
–0.00
–0.00
0.00
–0.14
Source: IMF staff calculations.
Note: The table shows the approximate differences between response paths of output growth, inflation, real interest rates, and real exchange
rates at three time horizons. The values shown in the first column reflect the total difference between the response paths in the high integration
version (2003 calibration) and response paths in the low integration version (1984 calibration). Values in the second to the fifth columns show
differences attributable to changes in each parameter group for all four blocks (holding all other parameters constant).
1Difference in average contract length.
151
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
minor role. This finding reflects two facts: first,
the calibrations allow for similar gains in flexibility and competition in all blocs, dampening the
relative effect of these changes in parameters
across the versions of the model; and second,
empirical evidence suggests that in fact mark-ups
and rigidities differ only marginally across the
two globalization environments (see also Table
3.8 in the main text).
References
Adler, Michael, and Bernard Dumas, 1983,
“International Portfolio Choice and Corporation
Finance: A Synthesis,” Journal of Finance, Vol. 38,
No. 3, pp. 925–84.
Angeloni, Ignazio, and others, 2004, “Inflation
Persistence in the Euro Area,” presented at the conference on Inflation Persistence in the Euro Area,
Frankfurt, December 10–11. Available via the
Internet: http://www.ecb.int/events/conferences/
html/inflationpersistence.en.html.
Bailliu, Jeannine, and Eiji Fujii, 2004, “Exchange-Rate
Pass-Through and the Inflation Environment in
Industrialized Countries: An Empirical Investigation,” Bank of Canada Working Paper No. 2004–21
(Ottawa: Bank of Canada).
Barba Navaretti, Giorgio, and Anthony J. Venables,
2004, Multinational Firms in the World Economy
(Princeton, New Jersey: Princeton University
Press).
Barnea, Ami, and Joseph Djivre, 2004, “Changes in
Monetary and Exchange Rate Policies and the
Transmission Mechanism in Israel, 1989.IV–2002.I,”
Bank of Israel Discussion Paper No. 2004.13
(Jerusalem: Bank of Israel).
Batini, Nicoletta, 2002, “Euro Area Inflation
Persistence,” ECB Working Paper No. 201
(Frankfurt: European Central Bank).
———, and Andrew G. Haldane, 1999, “ForwardLooking Rules for Monetary Policy,” in Monetary
Policy Rules, ed. by J.B. Taylor (Chicago, Illinois:
University of Chicago Press).
Batini, Nicoletta, Brian Jackson, and Stephen Nickell,
2002, “The Pricing Behaviour of UK Firms,” Bank
of England External MPC Unit Discussion Paper
No. 9 (London: Bank of England).
Batini, Nicoletta, and Edward Nelson, 2000, “When
the Bubble Bursts: Monetary Policy Rules and
Exchange Rate Market Behavior” (unpublished;
152
Washington and St. Louis, Missouri: IMF and
Federal Reserve Bank of St. Louis).
Batini, Nicoletta, and Kenny Turnbull, 2002, “A
Dynamic Monetary Condition Index for the UK,”
Journal of Policy Modeling, Vol. 24 (June), pp. 257–81.
Baxter, Marianne, and Urban J. Jermann, 1997, “The
International Diversification Puzzle Is Worse Than
You Think,” American Economic Review, Vol. 87
(March), pp. 70–80.
———, and Robert G. King, 1998, “Nontraded Goods,
Nontraded Factors, and International Non-diversification,” Journal of International Economics, Vol. 44
(April), pp. 211–29.
Bayoumi, Tamim, 2004, GEM: A New International
Macroeconomic Model, IMF Occasional Paper No. 239
(Washington: International Monetary Fund).
———, and others, 2004, “Exchange Rate Regimes,
International Linkages, and the Macroeconomic
Performance of the New Member States,” paper
presented at the IMF Jacques Polak Annual
Research Conference, November 4–5 (Washington:
International Monetary Fund).
Bertaut, Carol C., and Linda S. Kole, 2004, “What
Makes Investors Over or Underweight? Explaining
International Appetites for Foreign Equities,” FRB
International Finance Discussion Paper No. 819
(Washington: Board of Governors of the Federal
Reserve System).
Boivin, Jean, and Mark Giannoni, 2002, “Has Monetary
Policy Become Less Powerful?” Federal Reserve
Bank of New York Staff Report No. 144 (New York:
Federal Reserve Bank of New York). Available via
the Internet: http://www.newyorkfed.org/research/
staff reports/sr144.html.
———, 2003, “Has Monetary Policy Become More
Effective?” NBER Working Paper No. 9459
(Cambridge, Massachusetts: National Bureau of
Economic Research).
Bordo, Michael D., 1984, “The Gold Standard: The
Traditional Approach,” in A Retrospective on the
Classical Gold-Standard, 1821–1931, ed. by Michael
D. Bordo and Anna J. Schwartz (Chicago, Illinois:
University of Chicago Press).
———, and Barry Eichengreen, 1999, “Is Our
International Economic Environment Unusually
Crisis Prone?” in International Financial System
Conference Proceedings: Capital Flows and the Reserve
Bank of Australia, ed. by David Gruen and Luke
Gower (Sydney: Reserve Bank of Australia).
———, and Douglas A. Irwin, 1999, “Is Globalization
Today Really Different from Globalization a
REFERENCES
Hundred Years Ago? in Brookings Trade Forum, ed. by
Susan Collins and Robert Lawrence (Washington:
Brookings Institution Press).
Bordo, Michael D., Barry Eichengreen, and Jongwoo
Kim, 1998, “Was There Really an Earlier Period of
International Financial Integration Comparable to
Today?” in Conference Proceedings: The Implications of
the Globalization of World Financial Markets (Seoul,
South Korea: Bank of Korea), pp. 27–75.
Bordo, Michael D., and Hugh Rockoff, 1996, “The
Gold Standard as a ‘Good Housekeeping Seal of
Approval,’” Journal of Economic History, Vol. 56
(June), pp. 389–428.
Bordo, Michael, and others, 2001, “Is the Crisis
Problem Growing More Severe?” Economic Policy,
Vol. 16 (April), pp. 51–82.
Branson, William H., and Dale W. Henderson, 1984,
“The Specification and Influence of Asset Markets,”
in Handbook of International Economics, Vol. 2, ed. by
Ronald W. Jones and Peter B. Kenen (Amsterdam:
North-Holland), pp. 749–805.
Brash, Donald T., 2001, “Central Banks: What They
Can and Cannot Do,” speech to the Trans-Tasman
Business Circle, Sydney, Australia, March.
Bravo-Ortega, Claudio, and Julian di Giovanni, 2005,
“Trade Costs and Real Exchange Rate Volatility:
The Role of Ricardian Comparative Advantage,”
IMF Working Paper 05/5 (Washington:
International Monetary Fund).
Burstein, Ariel, Christopher J. Kurz, and Linda Tesar,
2004, “Trade, Production Sharing, and the International Transmission of Business Cycles” (unpublished;
New York: Federal Reserve Bank of New York).
Busse, Matthias, 2003, “Tariffs, Transport Costs, and
the WTO Doha Round: The Case of Developing
Countries,” Journal of International Law and Trade
Policy, Vol. 4, No. 1, pp. 15–31.
Calvo, Guillermo, 1983, “Staggered Prices in a UtilityMaximizing Framework,” Journal of Monetary
Economics, Vol. 12 (September), pp. 383–98.
———, and Carmen Reinhart, 2002, “Fear of
Floating,” Quarterly Journal of Economics, Vol. 117
(May), pp. 379–408.
Campa, José Manuel, and Linda Goldberg, 2002,
“Exchange Rate Pass-Through into Import Prices,”
NBER Working Paper No. 8934 (Cambridge,
Massachusetts: National Bureau of Economic
Research).
Chen, Hogan, and Kei-Mu Yi, 2003, “Vertical
Specialization and Trends in U.S. International
Trade” (unpublished).
Choudhri, Ehsan U., and Dalia S. Hakura, 2001,
“Exchange Rate Pass-Through to Domestic Prices:
Does the Inflationary Environment Matter?” IMF
Working Paper 01/194 (Washington: International
Monetary Fund).
Coenen, Günther, and Andrew Levin, 2004,
“Identifying the Influences of Nominal and Real
Rigidities in Aggregate Price-Setting Behavior,” ECB
Working Paper No. 418 (Frankfurt: European
Central Bank).
Cooper, Richard N., 2001, “Is the U.S. Current
Account Deficit Sustainable? Will It Be Sustained?”
Brookings Papers on Economic Activity: 1, Brookings
Institution, pp. 217–26.
———, 2004, “How Big Is the Hole in the Economy?”
Financial Times, November 1, p. 15.
Corsetti, Giancarlo, and Panagiotis Konstantinou,
2004, “Current Account Theory and the Dynamics
of U.S. Net Foreign Liabilities” (unpublished;
Rome: University of Rome).
Devereux, Michael B., Charles Engel, and Peter E.
Storgaard, 2004, “Endogenous Exchange Rate PassThrough When Nominal Prices Are Set in
Advance,” Journal of International Economics, Vol. 63
(June), pp. 263–91.
Devereux, Michael B., and James Yetman, 2002,
“Menu Costs and the Long-Run Output-Inflation
Trade-Off,” Economics Letters, Vol. 76 (June),
pp. 95–100.
Edey, Malcolm, and Ketil Hviding, 1995, “An
Assessment of Financial Reform in OECD
Countries,” Economics Department Working Paper
No. 154 (Paris: Organization for Economic
Cooperation and Development).
Edison, Hali, and others, 2002, “International
Financial Integration and Economic Growth,”
Journal of International Money and Finance, Vol. 21
(November), pp. 749–76.
Engel, Charles, and Akito Matsumoto, 2004, “Home
Bias in Equities Under New Open Economy
Macroeconomics,” presented at the Joint ECB/IMF
Workshop, Washington, November.
Faruqee, Hamid, Shujing Li, and Isabel K. Yan, 2004,
“The Determinants of International Portfolio
Holdings and Home Bias,” IMF Working Paper
04/34 (Washington: International Monetary
Fund).
Fischer, Stanley, 1997, “Capital Account Liberalization
and the Role of the IMF,” speech made at the IMF
Annual Meetings seminar on “Asia and the IMF,”
September 19, 1997. Available via the Internet:
153
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
http://www.imf/org/external/np/speeches/1997/
091997.htm.
———, 2001, “Exchange Rate Regimes: Is the Bipolar
View Correct?” Journal of Economic Perspectives,
Vol. 15 (Spring), pp. 3–24.
French, Kenneth, R., and James M. Poterba, 1991,
“Investor Diversification and International Equity
Markets,” American Economic Review, Vol. 81, No. 2,
pp. 222–26.
Gagnon, Joseph E., and Jane Ihrig, 2004, “Monetary
Policy and Exchange Rate Pass-Through,”
International Journal of Finance and Economics, Vol. 9
(October), pp. 315–38.
Galí, Jordi, Mark Gertler, and David López-Salido,
2001, “European Inflation Dynamics,” European
Economic Review, Vol. 45, No. 7, pp. 1237–70.
·
Ghironi, Fabio, Talan B. Is¸an, and Alessandro Rebucci,
2005, “Net Foreign Asset Positions and
Consumption Dynamics in the International
Economy,” IMF Working Paper (Washington:
International Monetary Fund, forthcoming).
Gonzalo, Jesús, and Clive W.J. Granger, 1995,
Estimation of Common Long-Memory Components
in Cointegrated Systems,” Journal of Business and
Economic Statistics, Vol. 13, No. 1, pp. 27–35.
Gonzalo, Jesús, and Serena Ng, 2001, “A Systematic
Framework for Analyzing the Dynamic Effects of
Permanent and Transitory Shocks,” Journal of
Economic Dynamics and Control, Vol. 25, No. 10,
pp. 1527–46.
Gourinchas, Pierre-Olivier, and Hélène Rey, 2004,
“International Financial Adjustment” (unpublished;
Princeton, New Jersey: Princeton University).
Greenspan, Alan, 2004, “Globalization and Innovation,” remarks at the Conference on Bank Structure
and Competition, sponsored by the Federal Reserve
Bank of Chicago, Chicago, Illinois, May 6.
Hanson, Gordon H., Raymond J. Mataloni, and
Matthew J. Slaughter, 2003, “Vertical Production
Networks in Multinational Firms,” NBER Working
Paper No. 9723 (Cambridge, Massachusetts:
National Bureau of Economic Research).
Harvey, Campbell R., 1991, “The World Price of
Covariance Risk,” Journal of Finance, Vol. 46
(March), pp. 111–57.
Heathcote, Jonathan, and Fabrizio Perri, 2004,
“Financial Globalization and Real Regionalization,”
Journal of Economic Theory, Vol. 119 (November),
pp. 207–43.
Huizinga, Harry, and Dantao Zhu, 2004, “Domestic
and International Finance: How Do They Affect
154
Consumption Smoothing?” CEPR Discussion Paper
No. 4677 (London: Centre for Economic Policy
Research).
Hunt, Benjamin, and Alessandro Rebucci, 2003,
“The U.S. Dollar and the Trade Deficit: What
Accounts for the Late 1990s?” IMF Working Paper
03/194 (Washington: International Monetary
Fund).
Imbs, Jean, 2004, “The Real Effects of Financial
Integration,” CEPR Discussion Paper No. 4335
(London: Centre for Economic Policy Research).
International Monetary Fund, 1993, Balance of
Payments Statistics Manual, Fifth Edition
(Washington).
Khan, Hashmat, 2004, “Price-Setting Behaviour,
Competition, and Mark-Up Shocks in the New
Keynesian Model,” Bank of England Working Paper
No. 240 (London: Bank of England).
Khan, Mohsin S., and Abdelhak S. Senhadji, 2000,
“Threshold Effects in the Relationship Between
Inflation and Growth,” IMF Working Paper 00/110
(Washington: International Monetary Fund).
King, Robert G., and others, 1991, “Stochastic Trends
and Economic Fluctuations,” American Economic
Review, Vol. 81, No. 4, pp. 819–40.
Lane, Philip, R., and Gian Maria Milesi-Ferretti, 2001,
“The External Wealth of Nations: Measures of
Foreign Assets and Liabilities for Industrial and
Developing Countries,” Journal of International
Economics, Vol. 55, No. 2, pp. 263–94.
———, 2002a, “External Wealth, the Trade Balance,
and the Real Exchange Rate,” European Economic
Review, Vol. 46 (June), pp. 1049–71.
———, 2002b, “Long-Term Capital Movements,” in
NBER Macroeconomics Annual, Vol. 16, ed. by Ben S.
Bernanke and Kenneth Rogoff (Cambridge,
Massachusetts: MIT Press), pp. 73–116.
———, 2003, “International Financial Integration,”
IMF Staff Papers, Vol. 50, Special Issue, pp. 82–113.
———, 2004a, “International Investment Patterns,”
IMF Working Paper 04/134 (Washington:
International Monetary Fund).
———, 2004b, “The Transfer Problem Revisited:
Net Foreign Assets and Real Exchange Rates,”
Review of Economics and Statistics, Vol. 86, No. 4,
pp. 841–57.
———, 2005a, “Financial Globalization and Exchange
Rates,” IMF Working Paper 05/3 (Washington:
International Monetary Fund).
———, 2005b, “The External Wealth of Nations Mark
II: Revised and Extended Estimates of Foreign
REFERENCES
Assets and Liabilities, 1970–2003” (unpublished;
Washington: International Monetary Fund).
Laxton, Douglas, and Paolo Pesenti, 2003, “Monetary
Rules for Small, Open, Emerging Economies,”
Journal of Monetary Economics, Vol. 50, No. 5,
pp. 1109–46.
Lettau, Martin, and Sydney Ludvigson, 2004,
“Understanding Trend and Cycle in Asset Values:
Reevaluating the Wealth Effect on Consumption,”
American Economic Review, Vol. 94, No. 1,
pp. 276–99.
Martins, Joaquim Oliveira, and Stefano Scarpetta,
1999, “The Levels and Cyclical Behaviour of MarkUps Across Countries and Market Structures,”
OECD Economics Department Working Paper No.
213 (Paris: Organization for Economic Cooperation
and Development).
Obstfeld, Maurice, 2002, “Exchange Rates and
Adjustment: Perspectives from the New OpenEconomy Macroeconomics,” Bank of Japan Monetary
and Economic Studies, Vol. 20 (December).
———, 2004, “External Adjustment,” NBER Working
Paper No. W10843 (Cambridge, Massachusetts:
National Bureau of Economic Research).
———, and Kenneth S. Rogoff, 1996, Foundations of
International Macroeconomics (Cambridge, Massachusetts: MIT Press).
———, 2000, “Perspectives on OECD Economic
Integration: Implications for U.S. Current Account
Adjustment,” in Global Economic Integration:
Opportunities and Challenges (Kansas City, Missouri:
Federal Reserve Bank of Kansas City).
———, 2001, “The Six Major Puzzles in International
Macroeconomics: Is There a Common Cause?” in
NBER Macroeconomics Annual 2000, Vol. 15, ed. by
Ben S. Bernanke and Kenneth Rogoff (Cambridge,
Massachusetts: MIT Press), pp. 338–90.
———, 2004, “The Unsustainable US Current
Account Position Revisited,” NBER Working Paper
No. 10869 (Cambridge, Massachusetts: National
Bureau of Economic Research).
Obstfeld, Maurice, and Alan M. Taylor, 2004, Global
Capital Markets: Integration, Crisis, and Growth (Cambridge, Massachusetts: Cambridge University Press).
Organization for Economic Cooperation and
Development, 2002, “Intra-Industry and Intra-Firm
Trade and the Internationalization of Production,”
Economic Outlook No. 71 (Paris), pp. 159–70.
Orphanides, Athanasios, 2003, “Historical Monetary
Analysis and the Taylor Rule,” Journal of Monetary
Economics, Vol. 50, No. 5, pp. 983–1022.
———, 2004, “Monetary Policy Rules, Macroeconomic
Stability, and Inflation: A View from the Trenches,”
Journal of Money, Credit and Banking, Vol. 36 (May),
pp. 151–75.
Pagan, Adrian, 2003, “Report on Modelling and
Forecasting at the Bank of England” (London:
Bank of England).
Panigirtzoglou, Nikolaos, 2004, “Implied Foreign
Exchange Risk Premia,” European Financial
Management, Vol. 10 (June), pp. 321–38.
Pesaran, M. Hashem, and Ron Smith, 1995,
“Estimating Long-Run Relationships from Dynamic
Heterogeneous Panels,” Journal of Econometrics,
Vol. 68, No. 1, pp. 79–113.
Prasad, Eswar S., and others, 2003, Effects on Financial
Globalization on Developing Countries: Some Empirical
Evidence, IMF Occasional Paper No. 220
(Washington: International Monetary Fund).
Rebucci, Alessandro, 2003, “On the Heterogeneity
Bias of Pooled Estimators in Stationary VAR
Specifications,” IMF Working Paper 03/73
(Washington: International Monetary Fund).
Rodrik, Dani, 2001, “The Developing Countries’
Hazardous Obsession with Global Integration,”
Kennedy School of Government, Harvard
University (Cambridge, Massachusetts: Harvard
University).
Rogoff, Kenneth, 2003, “Globalization and Global
Disinflation,” paper prepared for the Federal
Reserve Bank of Kansas City conference on Monetary Policy and Uncertainty: Adapting to a Changing
Economy, Jackson Hole, Wyoming, August.
Ruhl, Kim J., 2003, “Solving the Elasticity Puzzle in
International Economics” (unpublished;
Minneapolis, Minnesota: University of Minnesota).
Sargent, Thomas, 1999, The Conquest of American
Inflation (Princeton: New Jersey: Princeton
University Press).
Sbordone, Argia, 2005, “Do Expected Marginal Costs
Drive Inflation Dynamics?” Journal of Monetary
Economics (forthcoming).
Sinn, Stefan, 1990, “Net External Asset Positions of
145 Countries,” Kieler Studien No. 224, Institut für
Weltwirtschaft an der Universität Kiel (Tübingen,
Germany: J.C.B. Mohr).
Solnik, Bruno, 1974, “Why Not Diversify Internationally Rather Than Domestically?” Financial Analyst
Journal, Vol. 30 (July–August), pp. 48–54.
Stevens, Glenn, 2004, “Recent Issues for the Conduct
of Monetary Policy,” speech by the Reserve Bank of
Australia Deputy Governor, February 17.
155
CHAPTER III
GLOBALIZATION AND EXTERNAL IMBALANCES
Stiglitz, Joseph, 2000, “Capital Account Liberalization,
Economic Growth, and Instability,” World
Development, Vol. 28, No. 5, pp. 1075–86.
Swiss National Bank, 2003, “Study Center Gerzensee
Conference on Monetary Policy Under Incomplete
Information (Gerzensee, Switzerland, October 12,
2000),” Journal of Monetary Economics, Vol. 50
(April), pp. 497–720.
Taylor, John B., 2000, “Low Inflation, Pass-Through, and
the Pricing Power of Firms,” European Economic Review
(The Netherlands), Vol. 44 (June), pp. 1389–408.
Tesar, Linda L., 1993, “International Risk Sharing and
Non-Traded Goods,” Journal of International
Economics, Vol. 35 (August), pp. 69–89.
———, and Ingrid M. Werner, 1995, “Home Bias and
High Turnover,” Journal of International Money and
Finance, Vol. 14 (August), pp. 467–92.
Tille, Cedric, 2003, “The Impact of Exchange Rate
Movements on U.S. Foreign Debt,” Federal Reserve
Bank of New York, Current Issues in Economics and
Finance, Vol. 9 (January) pp. 1–7.
156
———, 2004, “Financial Integration and the Wealth
Effect of Exchange Rate Fluctuations” (unpublished; New York: Federal Reserve Bank of New
York).
Tytell, Irina, and Shang-Jin Wei, 2004, “Does Financial
Globalization Induce Better Macroeconomic
Policies?” paper presented at the Fifth Annual IMF
Research Conference, Washington, November 4.
Warne, Anders, 1993, “A Common Trends Model:
Identification, Estimation, and Inference,”
International Economic Studies Paper No. 555
(Stockholm: University of Stockholm, Institute for
International Economic Studies).
Williamson, John, and Molly Mahar, 1998, “A Survey
of Financial Liberalization,” Essays in International
Finance No. 211 (Princeton, New Jersey: Princeton
University, Department of Economics, International
Economics Section).
Yi, Kei-Mu, 2003, “Can Vertical Specialization Explain
the Growth of World Trade?” Journal of Political
Economy, Vol. 111, No. 1, pp. 52–102.